<%BANNER%>

Image Guidance Methods in Deep Brain Stimulation

Permanent Link: http://ufdc.ufl.edu/UFE0041600/00001

Material Information

Title: Image Guidance Methods in Deep Brain Stimulation
Physical Description: 1 online resource (111 p.)
Language: english
Creator: Sudhyadhom, Atchar
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: dbs, image, mri
Nuclear and Radiological Engineering -- Dissertations, Academic -- UF
Genre: Nuclear Engineering Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Deep Brain Stimulation (DBS) has shown promise as an alternative therapy for medication refractory neurological disorders (such as Parkinson s disease, essential tremor, and dystonia). DBS requires millimeter accuracy in the targeting of specific deep brain structures. Unfortunately, standard imaging methods (CT, T1 and T2-weighted MRI) have not previously shown significant anatomic contrast of structures that are targeted in DBS. In order to enhance surgical targeting in DBS, we have developed several tools to aid in both indirect and direct targeting of subcortical brain regions. The tools that we have created include the application of a deformable brain atlas, a novel MRI scan (the Fast Gray Matter Acquisition T1 Inversion Recovery, FGATIR) for differentiation of subcortical structures, application of diffusion tractography for localization of functional subregions of the brain, and a complete clinical platform to provide image guidance integrating all the tools mentioned previously. The methods developed in this work were evaluated to determine their potential clinical utility. The FGATIR scan was found to have enhanced contrast of three of the most common structures targeted for DBS versus standard scans. To determine potential targeting advantages of the FGATIR scan, a study was conducted that compared different targeting methods on the STN. This study revealed that targeting using the FGATIR was not significantly better than more basic methods of targeting for the specific target studied, the STN. A potential reason that more advanced imaging methods are not significantly more useful is that the STN is a target that may be consistently targeted based off known distances from landmarks alone. A study was conducted to compare probabilistic diffusion tractography (PDT) delineated regions for sensory and motor within the thalamus. These PDT delineated regions showed correlation with sensory and motor atlas and microelectrode data. A software and hardware system was developed that integrated all of these methods (as well as standard imaging methods) into a single multi-modal platform. This system allows for clinical targeting based off of all these methods combined together. The end result of this work is a clinically feasible methodology for multi-modal targeting of DBS integrating multiple novel techniques.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Atchar Sudhyadhom.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Bova, Frank J.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0041600:00001

Permanent Link: http://ufdc.ufl.edu/UFE0041600/00001

Material Information

Title: Image Guidance Methods in Deep Brain Stimulation
Physical Description: 1 online resource (111 p.)
Language: english
Creator: Sudhyadhom, Atchar
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2010

Subjects

Subjects / Keywords: dbs, image, mri
Nuclear and Radiological Engineering -- Dissertations, Academic -- UF
Genre: Nuclear Engineering Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Deep Brain Stimulation (DBS) has shown promise as an alternative therapy for medication refractory neurological disorders (such as Parkinson s disease, essential tremor, and dystonia). DBS requires millimeter accuracy in the targeting of specific deep brain structures. Unfortunately, standard imaging methods (CT, T1 and T2-weighted MRI) have not previously shown significant anatomic contrast of structures that are targeted in DBS. In order to enhance surgical targeting in DBS, we have developed several tools to aid in both indirect and direct targeting of subcortical brain regions. The tools that we have created include the application of a deformable brain atlas, a novel MRI scan (the Fast Gray Matter Acquisition T1 Inversion Recovery, FGATIR) for differentiation of subcortical structures, application of diffusion tractography for localization of functional subregions of the brain, and a complete clinical platform to provide image guidance integrating all the tools mentioned previously. The methods developed in this work were evaluated to determine their potential clinical utility. The FGATIR scan was found to have enhanced contrast of three of the most common structures targeted for DBS versus standard scans. To determine potential targeting advantages of the FGATIR scan, a study was conducted that compared different targeting methods on the STN. This study revealed that targeting using the FGATIR was not significantly better than more basic methods of targeting for the specific target studied, the STN. A potential reason that more advanced imaging methods are not significantly more useful is that the STN is a target that may be consistently targeted based off known distances from landmarks alone. A study was conducted to compare probabilistic diffusion tractography (PDT) delineated regions for sensory and motor within the thalamus. These PDT delineated regions showed correlation with sensory and motor atlas and microelectrode data. A software and hardware system was developed that integrated all of these methods (as well as standard imaging methods) into a single multi-modal platform. This system allows for clinical targeting based off of all these methods combined together. The end result of this work is a clinically feasible methodology for multi-modal targeting of DBS integrating multiple novel techniques.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by Atchar Sudhyadhom.
Thesis: Thesis (Ph.D.)--University of Florida, 2010.
Local: Adviser: Bova, Frank J.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2010
System ID: UFE0041600:00001


This item has the following downloads:


Full Text

PAGE 1

IMAGE GUIDANCE METHODS IN DEEP BRAIN STIMULATION By ATCHAR SUDHYADHOM A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORID A IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 2010 1

PAGE 2

2010 Atchar Sudhyadhom 2

PAGE 3

To my family and friends, whose support, encouragement and above all else love made this and everything else possible 3

PAGE 4

ACKNOWLEDGMENTS I would like to start by expressing my si ncere gratitude to my mentor, Dr. Frank Bova, for his endless support and guidance th roughout the years. While most people are only given direction and funding from their mentors, I was given a vision and everything I needed to pursue it from mine. I would like to thank Dr. Kelly Foote and Dr. Michael Okun who welcomed me into the worl d of DBS and introduced me to a problem that happily kept me up late nights and weekends as I worked to find a solution to it. I would also like to thank all the remain ing members of my committee: Dr. Stephen Blackband, Dr. David Gilland, and Dr. Davi d Hintenlang. Without their comments and critiques, both I and my work would have been worse off. I would like to acknowledge the members of the RadioSur gery/Biology lab who all have provided me with help of various forms throughout the years: Dr. Didier Rajon, Russ Moore, and Mu Yang. As with any endeavor, there are countless people who have stood in the background and made it all possible. Whil e all of my friends have supported me throughout the years, I would like to especially thank those that helped me through the toughest of those times: Joshua Aubuchon, Jason Craggs, Keith McGregor, and Alexander Singh-Alvarado. I wo uld also like to thank my parents and my sister who never stopped believing in me. Finally, I would like to acknowledge a ll of the people and all of the work that came bef ore me. If I have seen further it is only by standing on the shoulders of giants.Isaac Newton 4

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS ..................................................................................................4 LIST OF TABLES ............................................................................................................7 LIST OF FIGURES ..........................................................................................................8 LIST OF ABBREVIATIONS .............................................................................................9 ABSTRACT ...................................................................................................................12 CHAPTER 1 INTRODUC TION....................................................................................................14 2 BACKGRO UND...................................................................................................... 16 Deep Brain Stimul ation (D BS)................................................................................16 Neuroanatomy of Target s and Target ing................................................................20 Neuroanatomy and DBS ...................................................................................20 Localization of Target.......................................................................................25 Targeting Methods for D eep Brain Stim ulation.......................................................27 Imaging for Deep Brai n Stimulat ion........................................................................30 3 ENHANCING ANATOMIC CONTRAST OF SUBCORTICAL STRUCTURES........33 Aims........................................................................................................................33 Backgroun d.............................................................................................................33 Materials and Methods Contrast Study................................................................35 Subjec ts............................................................................................................35 Preoperative Imaging .......................................................................................36 Surgical Pr ocedure...........................................................................................36 Microelectrode Recordi ng and Regist ration......................................................38 Atlas Creation and Deformati on.......................................................................39 Analys is............................................................................................................42 Results Cont rast St udy........................................................................................43 Qualitative Analysis...........................................................................................43 Quantitative Analysis........................................................................................45 Discussion and Conclusions Contrast Study.......................................................45 Methods Targeting St udy.....................................................................................52 Subjec ts............................................................................................................52 Preoperative Imaging .......................................................................................53 Surgical Pr ocedure...........................................................................................53 Targeting Study Procedur e...............................................................................54 Analysis of Target Po ints..................................................................................56 5

PAGE 6

Results Targeting St udy.......................................................................................56 Discussion and Conclusions Targeting Study......................................................57 4 LOCALIZATION OF FUNCTIONAL TERRITORIES OF SUBCORTICAL STRUCTUR ES.......................................................................................................69 Aims........................................................................................................................69 Backgroun d.............................................................................................................69 Materials and Methods............................................................................................70 Subjec ts............................................................................................................70 Preoperative Imaging .......................................................................................71 Surgical Pr ocedure...........................................................................................71 Microelectrode Recordi ng and Regist ration......................................................72 Diffusion Data Processi ng................................................................................73 Result s....................................................................................................................75 Hard versus Soft Delineation of C onnectivity Defined Re gions (CDR).............75 Microelectrode Data versus Diffusion Tr actogr aphy.........................................76 Atlas Data versus Diffusion Trac tography ........................................................79 Discussion and C onclusi ons...................................................................................80 5 A SYSTEM FOR MULTI-MO DAL IMAGE GU IDANCE...........................................89 Aims........................................................................................................................89 Methods and Ma terials............................................................................................89 Softwar e...........................................................................................................89 Image Acquisition and Proce ssing...................................................................91 Result s....................................................................................................................91 Discussion and C onclusi ons...................................................................................92 6 OVERALL CONTRIBUTION S AND FUTURE WORK.............................................97 LIST OF REFE RENCES.............................................................................................100 BIOGRAPHICAL SKETCH ..........................................................................................111 6

PAGE 7

LIST OF TABLES Table page 3-1 Parameters used for t he scans in th is study.......................................................60 4-1 Mean differences between cent roids of vari ous regi ons.....................................84 7

PAGE 8

LIST OF FIGURES Figure page 2-1 T1-weighted MRI of the basal ganglia with atlas contours over laid....................32 3-1 Sagittal images of subcortical st ructur es............................................................61 3-2 Subcortical images from a ST N DBS patient (p re-surger y)................................62 3-3 Subcortical images from a thalamic Vim DBS patient (pre-surgery)...................63 3-4 Subcortical images from a GP i DBS patient (p re-surger y).................................64 3-5 Contrast ra tios ( CR)............................................................................................65 3-6 Simulated FGATIR images at various slice th icknesses.....................................66 3-7 Absolute errors in target planni ng using various ta rgeting me thods...................67 3-8 Corrected (for bias) absolute errors in target planning usi ng various targeting methods ..............................................................................................................68 4-1 Diagram of ROIs us ed in this study....................................................................85 4-2 Cumulative histogram of sensory and passive moto r responsive cells...............86 4-3 Exemplar cases of S1 and M1 tracts from the thal amus....................................87 4-4 Exemplar case of tracts seeded from t he thalamus to different portions of the M1 primary mo tor cort ex.....................................................................................88 5-1 Diffusion weighted secondary (b = 0) images fused to an FGATIR primary image ..................................................................................................................94 5-2 FGATIR primary images ov erlaid with atla s contour s.........................................95 5-3 FGATIR primary images overlaid with atlas contours and S1 and M1 tracks.....96 8

PAGE 9

LIST OF ABBREVIATIONS AC Anterior Commissure ALIC Anterior Limb of Internal Capsule BFMDRS Burke-Fahn-Marsden Dystonia Rating Scale CDR Connectivity Defined Region CM Centromedian Nucleus CNR Contrast to Noise Ratio CR Contrast Ratio CSF Cerebrospinal Fluid CT Computerized Tomography DBS Deep Brain Stimulation DTI Diffusion Tensor Imaging FGATIR Fast Gray Matter Acquisi tion T1 Inversion Recovery FLAIR Fluid Attenuated Inversion Recovery FSE Fast Spin Echo FSE-IR Fast Spin Echo Inversion Recovery GPe Globus Pallidus Externa GPi Globus Pallidus Interna HFS High Frequency Stimulation M1 Primary Motor Cortex MCP Mid-Comminssural Point MD Medial Dorsal MER Microelectrode Recording MFB Medial Forebrain Bundle MPTP 1-methyl-4-phenyl-1, 2,3,6-tetrahydropyridine 9

PAGE 10

MP-RAGE Magnetization Prepared-R apid Acquisition Gradient Echo MRI Magnetic Resonance Imaging NAC Nucleus Accumbens PC Posterior Commissure PD Parkinsons Disease PLIC Posterior Limb of Internal Capsule PPN Paramedian Pontine Nucleus ROI Region of Interest RSB Radiosurgery and Biology Lab S1 Primary Somatosensory Cortex SENSE Sensitivity Encoding SMA Supplementary Motor Area SNc Substantia Nigra Pars Compacta SNr Substantia Nigra Pars Reticulata STH Shands Teaching Hospital STIR Short Tau Inversion Recover STN Subthalamic Nucleus TE Time of Echo TI Inversion Time UPDRS Unified Parkinsons Disease Rating Scale Vc Ventralis Caudalis Nucleus Vim Ventralis Intermedius VL Ventral Lateral Voa Ventralis Oralis Anterior Vop Ventralis Oralis Posterior 10

PAGE 11

VPL Ventral Post erior Lateral VPM Ventral Posterior Medial VTA Ventral Tegmental Area ZI Zona Inserta 11

PAGE 12

Abstract of Dissertation Pr esented to the Graduate School of the University of Florida in Partial Fulf illment of the Requirements for t he Degree of Doctor of Philosophy IMAGE GUIDANCE METHODS IN DEEP BRAIN STIMULATION By Atchar Sudhyadhom May 2010 Chair: Frank J. Bova Major: Nuclear Engineering Sciences Deep Brain Stimulation (DBS) has shown promise as an alternative therapy for medication refractory neurolog ical disorders (such as Park insons disease, essential tremor, and dystonia). DBS require s millimeter accuracy in the targeting of specific deep brain structures. Unfortunately, standard imaging methods (CT, T1 and T2-weighted MRI) have not previously shown significant anatomic contrast of structures that are targeted in DBS. In order to enhance surgic al targeting in DBS, we have developed several tools to aid in both indirect and direct targeting of subcorti cal brain regions. The tools that we have created incl ude the application of a defo rmable brain atlas, a novel MRI scan (the Fast Gray Matter Acquisiti on T1 Inversion Recovery, FGATIR) for differentiation of subcortical structures application of diffusion tractography for localization of functional subregions of the brain, and a complete clinical platform to provide image guidance integrating all the tools ment ioned previously. The methods developed in this work were evaluated to determine their potential clinical utility. The FGATIR scan was found to have enhanced contrast of three of the most common structures targeted for D BS versus standard scans. To determine potential targeting advantages of the FGATIR scan, a study was conducted that 12

PAGE 13

compared different targeting methods on the STN. This study revealed that targeting using the FGATIR was not significantly bette r than more basic methods of targeting for the specific target studied, the STN. A potential reason that more advanced imaging methods are not significantly more useful is that the STN is a target that may be consistently targeted based off known distances from landmarks alone. A study was conducted to compare probabilistic diffusion tractography (PDT) delineated regions for sensory and motor within the thalamus. These PDT delineated regions showed correlation with sensory and motor atlas and microelectrode data. A software and hardware system was developed that integrated all of these methods (as well as standard imaging methods) into a single multi-modal platform This system allows for clinical targeting based off of all these me thods combined together. The end result of this work is a clinically f easible methodology for multi-modal targeting of DBS integrating multiple novel techniques. 13

PAGE 14

CHAPTER 1 INTRODUCTION In deep brain stimulation (DBS), the plac ement of the stimul ating electrode, requiring accuracy on the order of a millimeter has a significant impact on the efficacy of treatment. From the literature, there is no well est ablished position that an electrode can be implanted relative to anatomic landmarks. Previous groups have developed new anatomic imaging methods to be able to better delineate the structures of interest for DBS. Unfortunately, some of these met hods have not been shown to be clinically feasible, lack spatial resolution for accuracy, or have shown inconsistent results. In addition from neuroanatomy and intraoperativ ely obtained microelectrode recording (MER) maps, it is known that the structures of interest have mult iple circuits that comprise them. In the case of one of the most targeted structures, the subthalamic nucleus (STN), only one-third of the structure is part of the sensorimot or circuit targeted for DBS implantation. The region to be target ed is not necessarily anatomically distinct from its surrounding structures, so methods that can apply functional or physiological contrast from the neighboring structures ma y be able to better delineate the region for efficacious DBS stimulation. New interest in applying diffusion tensor imaging (DTI) tractography to neurosurgical applications ha s created some opportunities to use tract data for targeting and avoidance in neurosurgical procedures adding guidance information that has never before been acquir ed. Software and algorithms to analyze and integrate high resolution diffusion tract ography acquired for DBS do not currently exist and must be developed for clinical application and eval uation of accuracy. Using new image guidance methods devel oped in this work (FGATIR, diffusion tractography 14

PAGE 15

and targeting software), a clinically feas ible method was created and evaluated for usage in targeting of regions of interest for DBS. 15

PAGE 16

CHAPTER 2 BACKGROUND Deep Brain Stimulation (DBS) Deep brain stimulation (DBS) has becom e an attractive alternative for the treatment of several neurologic disorders. Indications for DBS (including clinical and research applications) now may include Parkinson disease (PD), essential tremor, dystonia, Tourettes syndrome, cluster headac he, epilepsy, depression, and obsessive compulsive disorder. Potential indications continue to expand. DBS has emerged as an attractive replacement for ablative therapies primarily due to reversibility and device programmability. Although medication is typi cally the primary treat ment for most basal ganglia disorders, a lack of efficacy, refracto riness to certain classes of medication, and delayed side effects have now opened the door to stereotaxic surgery as an alternative for select patients. Perhaps the most striking example of the phenomenon of failed medical therapy has been in the case of PD, where, although medications can be effective for the treatment of bradykinesia, rigidity, and tremor, many patients develop disabling dyskinesias following 5 or more years of medication therapy. The work of Benabid et al. (Benabid et al., 1989) has led to significant developments in chronic therapeutic brai n stimulation and has propelled DBS to become a highly efficacious and powerful tr eatment for PD and essential tremor. Historically, the thalamus and pallidum we re the most common targets for lesion therapy and DBS (Okun and Vitek, 2004). Howe ver, recent developments have led to the emergence of the subthalamic nucleus (STN) as an alter native target in movement disorders, and more recently, other target s such as the paramedian pontine nucleus, the hypothalamus, the nucleus accumbens, and t he anterior limb of the internal capsule 16

PAGE 17

(ALIC) have also emerged (Franzini et al., 2003; Limousin et al., 1998; Nuttin et al., 2003; Okun et al., 2004a). Although great stri des have been made in DBS research, the exact mechanism for its therapeutic effect remains unknown (McIntyre et al., 2004a). The mechanism may be partially due to a functional ablation of the target (an informational lesion) or alternatively, an inhibition of local cells with excitation of fibers of passage. Although the exact physiologic mec hanism has not been entirely elucidated, it is likely that a combination of effects acc ounts for the therapeutic ben efit, and there is a dependence on the anatomic location of neighbori ng structures for pr edicating positive and negative effects. The clin ical benefit derived from DBS may be further enhanced by careful placement (within mill imeters of accuracy) of the stimulating electrode and a significant investment in ti me and expertise in the adjustm ent of stimulation settings (Kumar, 2002; McClelland et al., 2005; Okun et al., 2005; Starr et al., 2002; Volkmann et al., 2002). Misplacement of leads can direct ly influence associative, limbic, motor, and/or surrounding regions and can lead to t he improvement or, in some cases, worsening, of symptoms. There are several theories regarding the exact mechanism of high-frequency stimulation (HFS). Although these theories are useful for explaining the effects seen in clinical and research studies, no single model has yet to explain the full spectrum of therapeutic effects. Whereas stimulation appears to offer similar benefits when compared with lesion therapies, the underlying mechanisms are believed to be different. Lesions modulate neural circuitry through the direct ablation of subnuclei, whereas stimulation likely inhibits neurons closest to the electrode while activating efferent fiber pathways. Current proposed mechanisms incl ude the following (Filali et al., 2004; Grill 17

PAGE 18

et al., 2004; McIntyre et al., 2004c): 1) synaptic inhibition through activation of inhibitory afferents (Dostrovsky et al., 2000); 2) synaptic depression through high-frequency interference of efferents (Hashimoto et al ., 2003); 3) depolarization blockade through the inactivation of voltagegated channels (Beurrier et al., 2001); 4) depletion of neurotransmitters leading to synaptic failu re; and 5) changes in downstream network activity as a result of stimulation. Further research into the neuroanatomy and neurophysiology of the basal ganglia will help to elucidate the circuits and pat hways involved in related diseases. In particular, 1-methyl-4-phenyl-1,2,3,6-tetr ahydropyridine (MPTP)-induced parkinsonism has allowed for the developm ent of animal models that have aided in the understanding of motor and non-motor loops of basal gang lia (Okun and Vitek, 2004). Changes in firing rates and patterns in the direct and indi rect basal ganglia pathways are thought to result in many of the symptoms of PD. These abnorma l brain signals can be neuromodulated by both lesioni ng and the application of el ectrical current through a DBS lead. The movement toward neuromodulation has led to recent and exciting therapeutic strategies for treatme nt of basal ganglia diseases. In current box models, there are two basic motor circuits of the basal ganglia, the indirect and direct pathways. Under normal circumstances and as the result of basal ganglia disease, disease firing rates and pa ttern are proposed to change. Target structures, represented by boxes, can by neuromodulated by lesions or DBS, which change output and have resultant upstream effects (Albin et al., 1989; Alexander et al., 1986). PD animal models have provided insight into the pathways involved in the normal and PD basal ganglia represented by these models. 18

PAGE 19

Alternatively, Mink (Mink, 1996) has prev iously described the basal ganglia as a system of two primary input st ructures, two primary output structures, and two intrinsic nuclei. Under his model, the striatum and ST N are considered to be input structures receiving excitatory input fr om the cerebral cortex and fr om the motor areas of the frontal lobe. The intrinsic nuclei include the GPe and the substantia nigra pars compacta (SNc). The GPe receives exci tatory and inhibitory input from the STN and striatum and provides inhibitory output to the STN, GPi, and SNr. The SNc, however, provides only an input and output loop with st riatum and serves as a center for dopamine-containing neurons. The output structures, GPi and SNr, both receive ex citatory input from STN and inhibitory input from striatum. They also provide output to motor regions of the brainstem as well as the ventral anterior ( VA) and ventral lateral (VL) nuclei of the thalamus. The inhibitory nat ure of GPi and SNr facilitates desired movements through surround inhibition of competing motor patterns in thalamocortical and brainstem circuits (Mink, 2003). Thus, under this model, diseases of the basal ganglia are ultimately the result of an inability to fa cilitate desired movements and/or inhibit competing motor patterns. This review addresses recent and impor tant research in the areas of neuroanatomy/neurophysiology as well as DBS clinical outcomes, with a particular emphasis on the importance of the sensorimotor, limbic, and associative circuits of the basal ganglia. Current models (Alexander et al., 1986) have proposed the existence of a family of segregated circuits and have proposed that within specific target structures there exist sensorimotor, limbi c, and associative regions. Previous work has shown that non-motor regions are in close proximity to motor areas. This proximity is important 19

PAGE 20

because activation of these neighboring regi ons may be responsible for non-motor side effects. These side effects highlight the importance of proper localization of the stimulating lead in order to maximize benefits and minimize adverse effects. Neuroanatomy of Targets and Targeting Traditional neuroanatomical atlases have prev iously been created by using myelin staining to reveal gray and white matter. Unfo rtunately, myelin stai ning will typically not be sufficient to distinguish certain neural elements, particularly when the primary difference is physiologic rather than anatomic Due to the difficulty involved with human in vivo studies of the functional anatomy of the basal ganglia, much of the current data are derived from primat e or rat studies. Recent studies (Karachi et al., 2002; Karachi et al., 2005) have focused on axonal tracing and immunoreactivity in primates. Calbindin immunoreactivity has been specifically show n to correspond with functional territories (Francois et al., 1994). Poor, intermediate, and strong immunoreactivity of calbindin can be correlated to sensorimotor, associative, and limbic regions, respectively. Additionally, imaging studies are becoming increasingly used to visualize the sensorimotor regions of basal ganglia (Lehericy et al., 2006). Neuroanatomy and DBS The most prominent nuclei of the basal ganglia include the striatum, globus pallidus, STN, and substantia nigra. The st ructures can be further subdivided into a ventral striatal region which may include caudate, putamen, and nucleus accumbens. The globus pallidus is divided into the inte rnal (GPi) and external segments (GPe). The substantia nigra is divided into two main divisions, the pars compacta (SNc) and the pars reticulata (SNr). The three typical targ ets of interest in DBS for PD include the ventralis intermedius nucleus of the thalamus (Vim), the GPi, and the STN. Generally in 20

PAGE 21

DBS we attempt to place the lead within the sensorimotor regions of these structures and avoid the limbic and associ ative regions (Figure 2-1). The Vim is a small structure, approximately 210 to 234 mm 3 (Butson and McIntyre, 2006), which is located in the ventral lateral tier of the thalamus (Figure 2-1). In the thalamus there are territorial areas within each subnuclei. The limbic circuit of basal ganglia originates from ventral pallidum and connects to the medial dorsal (MD) nucleus of the thalamus and ultimately to the medial prefrontal cortex (A lexander et al., 1990). The associative circuit has similar connections to the MD nucleus (Parent and Hazrati, 1995). The motor circuit, however, is derived from pallidum and substantia nigra projections into the ventralis oralis anterior (Voa) and ventralis oralis posterior (Vop), as well as the centromedian nucleus (CM) of th e thalamus (Alexander et al., 1990; Barbas et al., 1991; DeVito and Anderson, 1982; Francois et al., 1988; Parent and De Bellefeuille, 1982; Parent and Ha zrati, 1995), with final connecti ons to the frontal cortex. The specific divisions of the lateral thalam ic nuclei are of interest in DBS. The Vim nucleus serves as a cerebellar receiving area, and the Voa and Vop serve as pallidal receiving regions. In addition, the ventral pos terior medial (VPM) and ventral posterior lateral (VPL) nuclei of the thalamus compri se mainly somatosensory regions (Hirai and Jones, 1989). Due to the location of the somatos ensory territory in th e thalamus relative to associative and limbic territories, current fr om stimulation in these sensorimotor areas may not spread enough to creat e adverse cognitive, mood, and emotional effects, especially in patients implanted with Vi m DBS for tremor (Pahwa et al., 2001; Schuurman et al., 2000). However, the current usually causes internal capsule pulling, speech problems, and ataxia resulting from spread into surrounding structures and 21

PAGE 22

regions. More data is needed on newer targets su ch as the CM nucleus, which is now being employed for Tourettes syndrome (Visser-Vandewalle et al., 2003). The GPi is a relatively large, elliptica lly shaped structure that is 364 to 478 mm 3 (Karachi et al., 2002; Yelnik, 2002) and is lo cated posterior and ventral to the GPe and anterior to the posterior limb of the internal capsule (PLI C). Although the exact internal territory definitions of the GPi have not been as clearly demarcated anatomically as that of the thalamus, current evidence suggests t hat the sensorimotor territory dominates this structure, with the exis tence of a smaller associat ive region and an even smaller limbic territory (Karachi et al., 2002; Kelly and Strick, 2004; Middleton and Strick, 2000). Karachi et al. (Karachi et al., 2002) examin ed the volumes of human GPi sensorimotor, associative, and limbic territories through calb indin immunoreactivity, revealing relative proportions of 53%, 29%, and 18%, respectively The location of the sensorimotor region of the nucleus is in the postero-ventra l aspect of GPi, with associative and limbic regions comprising the antero-medial part of the nucleus. The motor portion of GPi is proportionally larger than the analogous region of the STN. This may explain why GPi DBS may exert less effect on specific mood and neuropsychological changes as a result from current spread into non-motor areas. Additional ly, there are fewer adjacent non-motor fiber bundles than in the STN r egion. Serotonergic, dopaminergic, and cholinergic fibers form rich pl exuses medial to GPi (approximately the central level in the rostrocaudal plane and only a few millimete rs from the sensor imotor area), which travel underneath to the ansa lenticularis en r oute to the internal and external medullary lamina. The fibers then run to the striatum and ultimately to the cerebral cortex. There is also a contingent of these fibers that pass dire ctly through the GPi to reach the striatum. 22

PAGE 23

Cells in the medial pole of GPi are related to limbic functions and project to the lateral habenula. Thus, stimulation through a media lly placed lead within the GPi may change the activity of these cells and lead to emotional and neuropsychol ogical changes. A more centrally placed GPi DBS electrode would be significantly less likely to affect the function of these areas. GPi DBS may also change activity in the nigro-striato-pallido-habenulo-n igro loop (by activation of the GPi/habenula projection) and lead to mood and cognitive changes. The lateral habenular nucleus receives afferent innervation from lim bic structures including the diagonal band, the lateral preoptico-hypothalamic area, t he substantia innominata, and the GPi. Additional limbic system inputs that may be affected by stim ulation are derived from the ventral tegmental area, the raphe nucleus, and the ventral periaqueductal grey. The STN is a biconvex structure, det ermined from our atlas data (Sudhyadhom, 2005) to be approximately 158 to 167 mm 3 (Yelnik, 2002), located ventral to the zona incerta (ZI) and dorsal to the SNr (Figure 2-1). The sensorimotor region of the nucleus is located posterior and dorsolaterally. The asso ciative region is localized dorsomedial in the nucleus, with the limbic region located at the medial tip (Car penter et al., 1981; Nakano et al., 1990; Nauta and Cole, 1978). Ther efore, ventral stim ulation will most likely affect cognitive circuits and medial stim ulation, mood-related circuits. Medial STN, the lateral hypothalamus, the medial forebrain bundle (MFB), and ZI all have connections with limbic and associative circuits and current spread to these structures from DBS electrodes may produce changes in mood and cognition. Stimulation in medial portions of the STN ma y spread current into the lateral hypothalamic zone. Most of the territory in this zone is occupied by a diffuse neuronal matrix that is known as the 23

PAGE 24

lateral hypothalamic area, as well as the bed nucleus of the MFB. The fibers of the MFB lie adjacent to the STN and may be activated by stimulation in more medial portions of the nucleus. The MFB are composed of loos ely arranged thin fibers connecting the septal area to the tegmentum of the midbrain. They pa ss through the medial parts of the mesencephalic and rhombenc ephalic tegmental areas and pass next to the raphe nuclei, whose fibers ascend to the hypothalamus and other limbic areas. The lateral fibers of the MFB sweep laterally and caudall y to the dorsal substantia nigra pars compacta, and then dorsomedially to centra l tegmentum and lateral tegmentum. They contain descending fibers from the amygdala and the bed nucleus of the stria terminalis, as well as connections to several hypothalamic areas. The desce nding fibers of the MFB project to several nuclei including the SNc, parabrachial nuclei, locus ceruleaus, nucleus subcoeruleus, noradrenergic areas A1, A2, A5, the superficial ventro-lateral reticular area, and the dorsal vagal comple x. The MFB also forms an ascending link between the forebrain and brainstem. Due to its diffuse connections within the limbic system, activation of the MFB may underlie and be causally related to the changes in mood and the development of cognitive deficit s associated with DBS in this region of the STN. In addition to medial STN, latera l hypothalamus, and the MFB, the dorsal zone of ZI, lying just above the STN, is also the target of limbic projections. The most caudal part of ZI is the peripeduncular nucleus, which has connections to the ventro-medial hypothalamus, amygdala, medial preoptic ar eas, and lateral hypothalamus. Stimulation in the region of ZI may, therefore, also result in changes in the limbic system. DBS in the region of the STN may additi onally activate fibers from brainstem ascending systems to prefrontal cortex and the thalamus. Fibers from the dorsal raphe 24

PAGE 25

nucleus, the SNc, the ventra l tegmental area (VTA), locu s coeruleus, and paramedian pontine nuclei (PPN) can also be affect ed because they follow the same general trajectory near the STN. These fiber systems a rrive at the level of STN to form a rich plexus medial to STN and then ascend along its dorsal surface (between STN and ZI) to reach the striatum and cortex. Localization of Target Although it is known that there are associative and lim bic territories within DBS target structures, it remains a challenge to accurately place the stimulating electrode in a position that maximizes benefit while avoi ding cognitive and behavi oral side effects. Additionally, side effects ma y occur from spread of current beyond the borders of the deep nuclei, particularly when surrounding stru ctures and fiber bundles are affected (MFB, hypothalamus, zona incerta, internal capsule, and so forth). As a consequence, the exact localization of electrodes or c ontacts has become an extremely important factor in the overall outcom e of the procedure. Several studies have examined these lead locations as potentially im portant predictors of outcome in PD and essential tremor patients (STN, GPi, and Vim targets). STN stimulation studies (Hamel et al., 2003; Hutchison et al., 1998; Lanotte et al., 2002; Nowinski et al., 2005; Saint-Cyr et al ., 2002; Starr et al., 1999) have shown varying locations for the clinically optimal lead location or contact, with areas ranging approximately 10 mm to 12 mm lateral, 2 mm to 6 mm inferior, and 1 mm to 4 mm posterior with reference to the mid-commissural point (MC P). Results of optimal lead localization from Saint-Cyr et al. (Saint -Cyr et al., 2002) showed that the lateral anterodorsal portion of STN had the best effica cy in PD. Localization of stimulating electrodes within the framework of a scalable atlas have also revealed a correlation of 25

PAGE 26

location with the optimal contact (Yelnik et al., 2003). The results of several groups indicate that optimal contact location may involve stimulation of surrounding structures (such as Fields of Forel and zona incerta), su ch as in the case of STN DBS (McIntyre et al., 2004b). Similar localization studies on pallidal stimulation do not seem to show as significant of a relationship with electr ode or optimal contact location and overall outcome. Starr et al (Starr et al., 2004) presented results that revealed leads that produced greater than 50% improvement in Burke-Fahn-Marsden Dystonia Rating Scale (BFMDRS) in patients with dystonia were located at approximately 19.3 mm lateral, 6.1 mm anterior, and 0.5 mm inferior to the MCP, but these locations were not statistically significant from lead locations that were suboptimal (this may have been a disease specific or dystonia subtype phenom enon). There are studies of lesion location for pallidotomy that may be useful in this regard. Although the typical region of stimulation (as well as Leksell pallidotomy) is the posterolateral portion of GPi, there have been previous studies that have shown motor benefit from stimulation of the anteromedial portion of GPi (Durif et al., 1999). Schrader et al. (Schrader et al., 2002) cites GPi stimulation standard coordinates of 21 to 22 mm lateral, 3 to 4 mm anterior, and 1 to 2 mm inferior to the MCP. For GPi, the effective volume that provides benefit for stimulation may be larger than that of other targets due to the proportionally larger motor region. This may account for the wider latitude in stimulation locations that provide benefit. The role of lead location and outcome in the pa llidum needs more careful study, particularly in light of the pallid otomy work by Gross et al. (Gross et al., 1999) showing that specific regi ons of the pallidum may affect symptoms in a differential 26

PAGE 27

way and may, therefore, be highly dependent on lead location. The optimal GPi stimulation site has not been well determi ned, and unfortunately t here are not enough studies in enough disorders addressed by DBS at this time to draw firm conclusions. Papavassiliou et al. (Papavassiliou et al., 2004) previously publ ished that tremor score in Vim DBS for essential tremor pati ents was correlated with lead location. Thirty seven patients data were analyzed in both DBS on and off states and lead locations were measured. The optimal electrode loca tion was determined to be 6.3 mm anterior to the posterior commissure and 10.0 mm lateral to the wall of the th ird ventricle. Other groups (Benabid et al., 1991; Koller et al ., 1997; Ondo et al., 1998) have found analogous correlations between electrode location and benefit. Targeting Methods for Deep Brain Stimulation Due to the inherent difficulty in determining the optimal clinical location of interest, the clinical procedure followed for DBS typica lly involves a multiple step process in which each step helps provide more information to localize the most efficacious location for stimulation. While each institution uses differing methods to localize the anatomy of interest, the procedure is typically done using image guidance met hods and often using preoperative imaging to provide an initial estimate. The DBS procedure followed at the University of Florida and Shands Teaching Hospital (STH) is an imaging based microelectrode guided procedure. In order to localize where to stimulate, multiple localization techniques must be utilized to determine the location of the target, which requires accuracy on the order of a millimet er. The targeting process involved can be divided into two major steps : 1) Imaging based targeting and 2) Microelectrode based targeting. 27

PAGE 28

The imaging based targeting can involve a number of imaging modalities. While typically tomographic imaging modalities su ch as computed tomography (CT) and magnetic resonance imaging (MRI) have been use to directly localize targets, imaging methods such as ventriculography, where a contrast medium is injected into the ventricles, can be used to visualize structur es around the target region. The procedure currently used at STH involves a two-step process for direct target ing imaging. In the first step, a pre-operative nonstereotactic MRI is acquired while in the second step a CT is acquired with a stereot actic headring on the patient. Fusion or image registration technology is then used to match the two im age sets together such that the MRI is localized to the space of the CT image set. The primary purpose of the multiple imaging methods is to allow for stereotactic tar geting from the MRI scan whose contrast characteristics are typically better suited for soft tissue. MRI scans te nd to offer higher soft tissue contrast and with certain types of scans such as T1-weighted images they can offer higher grey versus white matte r contrast than CT images. The secondary purpose is to potentially account for spatial inhomogeneities that ma y arise in the MRI acquisition process due to non-ideal magnetic field homogeneity. Since current imaging methods may not offer all the necessary contrast to be able to accurately localize the clinically appropria te region of stimulati on, additional targeting techniques are typically used that supplemen t the relatively poor image contrast and resolution. Another often used technique is to define a coordinate system based off anatomic points such as the anterior commissure (AC), the posterior commissure (PC), and a mid-line point that can be used to dete rmine anatomic location relative to known coordinates in an anatomic atlas. The structures of interest are anatomically located 28

PAGE 29

relative to this new coordinate system. Pr eviously, the radiosurgery and biology lab (RSB) have applied more extensive forms of atlas methods to help overcome some of the inherent limitations of this methodology with vary ing degrees of success. At STH, these coordinates are chosen and used to provi de an initial estimate of target location based off the modified Schaltenbrand-Bailey atlas (Sudhyadhom, 2005). Target location is further refined by linear deformation of an atlas to fi t with patient anatomy. While these refinements may help to define the anatomic location of interest, as shown by the previous review of targeting literature the anatomic location of interest may be difficult to identify and alone may not be sufficient to opt imally define the target. As a result, an additional map from microelec trode data is typically acquired intraoperatively to help better define the anatomic and func tional region for stimulation. Microelectrode recording (MER) maps ar e linear tracks of data that define anatomic and functional regions. These data are acquired by the implantation of a microelectrode in the region around the targ et. Typically an expert in interpreting and understanding the complex electrical signals is present during the operation to determine the location and type of cells that observed. This process provides the clinicians with an ability to create a map that is not limited by the anatomic and chemical differences present in the tissue but rather by the electrophysiology and the functionality of the cells in that region. Clinicians oft en use passive movement of limbs to provide additional information on the functional relations hip the cells have. Using this process, cells of a particular structure may be classified as belonging to a sensorimotor circuit which is key in localizing the active area fo r DBS. At STH, thes e linear MER maps are acquired at multiple positions to create a three-dimensional (3D) map that provides 29

PAGE 30

adequate information of the anatomy and physiology of the cells in region that was targeted. The combination of these maps with an anatomic atlas helps to provide a clearer picture of where in the patients anatomy the targeted regi on is compared to a standard atlas anatomy. While MER is c onsidered the standard in mapping for DBS implantation, it has the potential to increase the complic ation rate and can significantly increase the length of the procedure. Depending on the level of experience and mapping method employed, the quality of MER maps created may vary as the process of cell identification is somewhat subj ective. Once the mapping process has been completed, a location to best treat t he patients symptoms is determined and a stimulating electrode is implanted. Imaging for Deep Brain Stimulation The process of deep brain stimulation requires a significant emphasis on the quality of anatomic imaging and manipulation to be able to initially localize and accurately target the region of interest. As with any other stereotactic procedure, the imaging methods of the DBS procedure require high spatial resolution and high contrast of the anatomy to be targeted. While high spatial resolution (sub-millimeter) is often attainable using clinically feasible scan met hods, the contrast necessary to localize the region of interest for DBS has been elusive. With current CT technology, a whole brain scan is able to acquire a sub-millimeter in plane scan with 1 mm or less thick slices in less than a minute. Unfortunately, CT sc ans often have poor contrast between soft tissue structures. MRI scans often have the abili ty to increase the contrast of particular soft tissue structures that ar e beyond the contrast obtainable with CT scans, but at the expense of increased scanner time. 30

PAGE 31

The characteristics of the target structur es in DBS often pose some difficulty in imaging. Three of the most common targets in DBS are the Vim nucleus of the thalamus, STN, and GPi which ar e all small in size (within 5 mm in some dimensions) as mentioned previously. In addition to thei r small sizes, these structures boundaries have not been well defined through current ly viable imaging methods. Deoni e t al (Deoni et al., 2005b) has shown that thalamic nuclei can be resolved using T1-weighted imaging, but at clinically unfeasible scanner times of several hours. Similarly, other groups have shown the ability to resolve thalamic nuclei using diffusion tensor imaging but at resolution on the order of several m illimeters which may be too large to resolve thalamic nuclei of interest to DBS (Behr ens et al., 2003a; Wiegell et al., 2003). Other groups (Slavin et al., 2006) have proposed to delineate the STN using T2-weighted images, but it is not clear if the entire STN is visible using these methods in particular the lateral posterior portion that is of intere st in DBS stimulation. Similar optimizations for GPi have also been proposed using T1-w eighted inversion recovery sequences for enhanced contrast (Pinsker et al., 2008). Wh ile all these previous methods have been proposed as techniques to enhance the contrast of the target structur es, some of them employ clinically unacceptable scanner time s for the DBS patient population, lower resolution methods, or questionable contrast of the entire structure. In addition, none of these methods may solve the larger problem posed which is to identify the physiological or functional location corresponding to the clinical problem. 31

PAGE 32

A B Figure 2-1. T1-weighted MRI of the basal ganglia with atlas contours overlaid in A) sagittal and B) coronal views. The moto r region of GPi and STN are illustrated by shaded regions as given by somatot opic findings. Structures of the basal ganglia are marked by their respective names. 32

PAGE 33

CHAPTER 3 ENHANCING ANATOMIC CONTRAST OF SUBCORTICAL STRUCTURES Aims Our aims were to determine if the dev eloped high contrast T1-weighted FGATIR anatomic images had enhanced cont rast characteristics ve rsus standard T1-weighted MP-RAGE and T2-weighted FLAIR scans for the ventral lateral thalamus, subthalamic nucleus, and globus pallidus (specifically GP i) and to determine potential clinical utility of implementing the FGATIR s equence for targeting of STN. Background Deep brain stimulation (DBS) has become an accepted treatment for medication-refractory movement disorder s (DBSPDSG, 2001; Hung et al., 2007; Wider et al., 2008; Zorzi et al., 2005) and has also been employed for neuropsychiatric indications in several recent trials (Co syns et al., 2003; Greenberg et al., 2006; Lozano et al., 2008; Temel and Visser-Vandewalle, 2004). The procedure consists of placing a stimulating electrode into a specific brain stru cture with the intent of locally modulating a basal ganglia circuit, and consequently improvi ng clinical symptoms. The target chosen depends on the disorder being addressed and on the patients symptoms. Common targets have included the subt halamic nucleus (STN) (Wider et al., 2008), globus pallidus interna (GPi) (Hung et al., 2007), nucleus accumbens (NAc) and anterior limb of the internal capsule (ALIC) (Nuttin BJ et al., 2003; Okun et al ., 2007), cingulate cortex (area 25) (Lozano et al., 2008), and multiple thalamic subregions (Kumar et al., 2003; Pahwa et al., 2006). The stimulating electrode must be accurately placed within the target region for maximal efficacy (Chen et al., 2006); (Amirnovin et al., 20 06). Unfortunately, it may be 33

PAGE 34

difficult to image DBS targets precisely en ough to allow placement on the basis of stereotactic imaging alone. A number of ancillary methodol ogies have therefore been employed to improve the accuracy of electrode placement, including intraoperative microelectrode recording (MER), atlasbased mapping, and computer modeling (Chakravarty et al., 2008). These te chniques are limited by their dependence on operator experience and by the variation in brain anat omy between patients (Bootin, 2006; Chen et al., 2006; Duffner et al., 2002; Lee et al., 2005; Rampini et al., 2003). Both T1 and T2 weighted MRI have commonly been employed as adjuncts to stereotactic targeting. Previ ous groups have typically employed 2D fast spin echo (FSE) sequences to create these T2-weighted images. This method has significantly limited the ability of the operator to acquire sm all slice thickness images due to gradient hardware requirements. As a compromise m any DBS groups acquire thick slice images from multiple orientations (Dor mont et al., 2004; Kitajima et al., 2008; Reich et al., 2000; Slavin et al., 2006). T1-weighted imaging in the form of inversion recovery based 2D FSE sequences have also been used to delineate GPi (Pinsker et al., 2008; Reich et al., 2000) and thalamus (Deoni et al., 2005a; Me rcado et al., 2006), but these methods have been hampered by the need for very thick s lices or for longer scan times as well. Several groups (Kovacs et al., 2009; Lee et al., 2006; Pollo et al., 2007) have also used standard contrast high resolution ( 1 mm 3 ) T1-weighted MP-RAGE sequences for localization of where to implant the D BS electrode. These groups often use the T1w MP-RAGE sequence to localize landmarks (suc h as the AC and PC) rather than direct localization. In special cases other landmarks are used for indi rect localization, Lee et al (Lee et al., 2006) localized the STN on MP-RAGE scans through landmark 34

PAGE 35

measurements from the supramamillary co mmissure. A multi-landmark localization method using multiple imaging methods at 3T has also been previously used to locate the STN versus standard technique 1.5T imagi ng (Toda et al., 2009). In this case, the 3T technique using multiple localizati on techniques was found to enhance targeting versus the standard 1.5T imaging and targeting technique. Currently widely used clinical protocols have not previously produced both sufficient image contrast and resolution for DBS targeting based on MRI alone. We recently employed a 3T MRI scan protocol that improves upon standard high resolution 3T T1 and T2 protocols. This new approach, which will be referred to in this work as the Fast Grey Matter Acquisition T1 Inversion Recovery (FGATIR), seems in this pilot study to provide improved, high-re solution single-millimeter slice visualization of target structures with height ened grey/white matter contrast in regions of interest. In this chapter, we examine the contrast char acteristics of the FGATIR scan and compare it to previously used methods and then exam ine the effect of adding the FGATIR scan into the targeting process used at STH. Materials and Methods Contrast Study Subjects For this pilot study we examined the preoperative scans of three patients with advanced and medication refractory Parkinsons di sease (n = 2) or essential tremor (n = 1) who were to undergo DBS. These pati ents underwent formal evaluations by a fellowship-trained Movement Disorders Neur ologist, a Neurosurgeon, a Psychiatrist, and a Neuropsychologist in order to ensure accuracy of diagnosis by clinical criteria (Okun et al., 2004b) as well as absence of significant cognitive or psychiatric comorbidity. Prior to implantation, patients with PD were also required to demonstrate at 35

PAGE 36

least a 30% improvement in the motor subsection of the Unified Parkinsons Disease Rating Scale (UPDRS III) between the on and off medication states (Okun et al., 2004b). Preoperative Imaging Each included patient received four scans on the day prior to surgery: a three plane localizing scout, a T1-weighted 3D Magnetization PreparedRapid Acquisition Gradient Echo (MP-RAGE), a 3D T2-weight ed Fluid Attenuated Inversion Recovery (FLAIR), and a T1-weighted 3D FGATIR protocol (the last th ree scans were each single volume whole brain scans). All scans were acquired on a clinical Siemens Allegra 3T MRI using a quadrature birdcage headcoil. The total scanning time for all four scans was 30 minutes. Specific parameters and scan times for each scan utilized are listed in Table 3-1. The FGATIR protocol was developed fr om a standard clinical T1w MP-RAGE sequence by modifying the inversion pre-pulse timing. The inversion time (TI) was set to ~400 ms in order to nullify the white matter signal. The use of a short inversion time led to the contrast inversion of some regions re lative to standard T1; for example, normally dark cerebrospinal fluid (CSF) signal was bright. The echo time (TE) and other parameters were set such that the TI was th e dominant weighting factor in the contrast. Surgical Procedure On the morning of the operation, a Co smanRobertsWells (CRW) head ring was applied under local anesthes ia and a high-resolution stereotactic head computed tomography (CT) scan was performed. The CT and MRI images were fused using in-house computer software, a software package that is analogous to the Varian (Palo Alto, CA) system but with several added features that facilitated ta rgeting in atlas 36

PAGE 37

space by coregistering anatomical l andmarks. A Cartesian coordinate system confirmed the patients mid-commissural point and this point was used as a reference to confirm the target (Okun et al., 2007). The software utilized allowed the display and registration of a deformable 3D atlas (Sudhyadhom, 2005), based on the Schaltenbr and and Bailey stereotactic atlas (Schaltenbrand and Bailey, 1959) which was then overlaid onto the MRI scans. The atlas and scans were used together to target the structure of interest. We employed a two step targeting process. An initial esti mate based on anatomic (AC/PC/midline point) coordinates was then followed by direct ta rgeting adjustments using a deformed atlas overlay over image slices as well as direct visualization of target structures. Target selection depended on the underlying diseas e. Patients with ET had the anterior boundary of the ventralis caudalis nucleus (Vc) of the thalamus tar geted. This boundary was estimated based on indirect targeting and confirmed by the use of microelectrode recording (MER, detailed below). The final el ectrode location was 2 mm anterior to this point along the anterior boundary of the ventra l intermediate nucleus of the thalamus (Vim) and the ventralis oralis posterior (V op) nucleus. Those patie nts with PD had their initial target points within the motor (dorsolateral) STN or the motor (posterolateral) GPi. The indirect atlas coordinate for the STN tar get (tip of lead) for an initial MER pass was approximately AP -3 mm, LT 11 mm, and AX -7 mm. The indirect atlas coordinate used for the GPi target for an initial MER pass was approximat ely AP 1 mm, LT 21 mm, and AX -6 mm. Coordinates were modified from this indirect targeting by using atlas deformation and direct visualizat ion of target structures. 37

PAGE 38

Microelectrode Recording and Registration The target coordinates were verified to be wi thin the region of interest via multiple MER passes. Our technique used a 3D mapping procedure to guide electrode placement. For each pass, cellular activity was recorded at millimeter intervals beginning at 30 mm above the selected target, and at submillimeter intervals as the microelectrode approached the ta rget region. At each inte rval the encountered region was determined by the recording neurologist based on the sound and appearance of the recording and the depth at which it wa s observed. Each such determination was represented in real-time as a color-coded point overlaid on an individual patients MRI at the corresponding stereotactic coordinates. In addition to single cell recordings, cellular firing in response to passive motion and sens ory stimulation was used to delineate the somatotopic organization of the target structure. The use of registration software allowe d the translation of MER passes into a linear map of structure and somatotopy that was then overlaid on a patients MRI (as in Figures 3-2D, 3-3D, and 3-4D). A final decision, based on this aggregate map, was made as to the optimal location to place the permanent DBS elec trode. The procedure decision-making process took into account bot h electrode tip location within the target region and the electrodes proximit y to regions near the target that might result in side effects when stimulated. The precise mapping process used varied by target. For the three targets (STN, GPi, and Vim) we employed a true mapping strategy: we performed a single MER pass and used its results to determine the locati on of our next MER track. We typically performed three to five passes to confirm a targets boundaries. 38

PAGE 39

For STN and GPi, we first mapped the anterior-posterior plane intratarget somatotopic extremity and face responses We next located the boundary between the target structure and the internal capsule. This boundary was marked by a transition from the typical firing pattern of the target struct ure to the relative silence of the internal capsule. We then confirmed the lateral bor der of each structure (again by noting the MER transition from cellularity to relative silence). The inferior boundary is recorded near the termination of each ME R track by the transition to the tonic firing of SNr for the STN or, in the case of t he GPi, the transition to the li ght-sensitive activation of the fibers of the optic tract. For the STN we implanted 3 to 3.5 mm poste rior to the anterior border and 2.5 to 3 mm medial to t he internal capsule boundary with the deepest electrode contact placed at the STN/SNr boundar y. For GPi we typically implanted 2.5 to 3 mm anterior to the posteri or border (internal capsule) and 2 to 3 mm from the lateral border (GPe) with the deepest el ectrode contact placed immediately superior to the optic tract. In the case of Vim implantation the electr ode is inserted at a slightly shallower angle than that of the plane of t he typical Vc/Vim boundary. This allows us to locate the anterior border of Vc as we pass through the Vim side of the Vc/Vim border superiorly and the Vc side of that border inferiorly The transition from Vim to Vc can be appreciated as a change from motor-responsiv e to sensory-responsive cells. We typically implant the stimulating electrode 2 mm anterior to the hand region of Vc. Atlas Creation and Deformation The atlas used in this work is an in -house created atlas (S udhyadhom, 2005) that is based on the sagittal series contour s of the Schaltenbrand-Bailey atlas (Schaltenbrand and Bailey, 1959) of subcortical structures. Contours for each of these 39

PAGE 40

subcortical structures were created by approximating connecti ons between sagittal Schaltenbrand-Bailey atlas contours using De launay triangulation. By applying this algorithm across all sagittal contours, a linear surface with rough edges was produced which fitted points from one contour to corresponding points along neighboring contours. In order to provide a more continuous surface each of these structures was voxelized to a grid of 0.25 x 0.25 x 0.25 mm 3 These structures were then smoothed using a Gaussian filter of fu ll-width half maximum of 0.5 mm. This process reduced the discontinuities and abnormalities seen with the Schaltenbrand-Bailey atlas but still held relatively true to the original atlas contours. The final resu lt was a smooth surface that approximately matched the size, shape, and location of the original Schaltenbrand-Bailey atlas contours. This di gital atlas itself was manually validated against the original Schaltenbr and-Bailey atlas contours by overlaying the formers contours over the corresponding Sc haltenbrand-Bailey atlas planes. The resulting digital atlas can be displa yed and deformed to match the anatomy of each particular patient using nine linear degr ees of freedom: scaling, translation, and rotation in the medial-lateral, anterior-posterior, and superior-inferior axes. The general procedure for atlas alignment was 1) designat ion of an anterior/posterior commissure coordinate system (the coordi nate system by which the atlas was created in), 2) linear affine transformation of the atlas to a visua lly determined best fit to the patients anatomy, and 3) further fine adj ustments of the atlas to best fit the patients anatomy to the target region using nearby st ructures as reference points. In the case of the thalam us the boundaries are generally visible and can be used to make fine adjustments. For the STN, both the thalamus and substantia nigra 40

PAGE 41

boundaries (superior and inferior) can be used to make final adjustments. For the GPi, the medial boundary of striatum and the ant erior boundary of the anterior limb of the internal capsule are used for validation. Ea ch of these methods may be utilized with any of the three types of images examined in th is paper. Direct visualization of additional structures can be used to aid in the defo rmation in the cases of FLAIR and FGATIR imaging. We used two methods to evaluate the def ormation accuracy. First, a movement disorders trained neurosurgeon ve rified that atlas structur es neighboring the target region were aligned with imaging data. This neurosurgeon (who has performed approximately 500 DBS cases) completed the atlas deformation in all cases to match the patient anatomy with the MR imaging. All images (MPRAGE/FLAIR/FGATIR) were fused together to provide a complete set of data by which to complete the deformation. An AC/PC coordinate system was chosen fo r each patient which the atlas was then registered to, so that adjustments were made only after the atlas and patient anatomy were framed within the same coordinate syst em for all three (mutually fused) image sets. Second, MER maps were used to physiolog ically validate correlation between atlas and actual structures lying along the MER path. Intraoperative MER data for the patients in this study showed good corres pondence to atlas ROI locations (as seen in Figures 3-2D, 3-3D, and 3-4D), suggesting that the atlas is a reasonable fit around the targets of interest. 41

PAGE 42

Using both an experts evaluation and ME R data confirmation allowed us to increase our confidence that the atlas deformation chosen provided a robust fit to the structures around the targets. Analysis The T1-weighted, T2-weighted FLAIR, and FGATIR sequences were fused to ensure anatomical and stereotactic co-localization using in-hous e created software. Additional software was then ut ilized for the linear affine transformation of the 3D atlas to fit the patients anatomy (as was described previously). The atlas was then used to delineate structures of interest for qualitative and quantitat ive analysis. We qualitatively compared our ability to delineate boundaries for the regions of interest among the scan types employed, and quantitatively evaluated t he contrast to noise ratio and contrast ratio between areas believed to be within regi ons of interest and the surrounding area. MER data was used to electrophysiologically verify the boundaries of structures of interest. For this quantitative analysis, the ROI for ve ntral lateral (VL) thalamus, STN, and GPi were created by using the results of the registration of the defo rmable atlas. In the case of the VL thalamus, contrast ratios were calculated between the VL thalamus and the non-VL thalamus, as well as between the VL thalamus and the posterior limb of the internal capsule (PLIC). For t he STN, contrast ratios were calculated between the STN and the substantia nigra reticulata (SNr) as well as between the STN and the thalamus and finally between STN and ZI. For the GPi, contrast ratios were calculated between the GPi and the PLIC. These ratios are summ arized in Figure 3-5. The formulas used for contrast to noise ratio and contra st ratio are defined as the following: 42

PAGE 43

Contrast to Noise Ratio (CNR): |(S A S B )|/(Standard deviation of background noise) Contrast Ratio (CR): |(S A S B)|/(S B A ) S A and S B are average intensity values for regions A and B, respectively, as determined from ROI created fr om atlas delineated regions. A set of the voxels outside the patients head were taken to sample the background noise and this region was used to determine the standard deviation of backg round noise used in the CNR calculation. This technique is a commonly used technique for estimation of the background noise in a MR image. The quantitative measur e of contrast to noise rati o provides an estimate of the contrast expected between two regions versus the background noise (which would be expected to inhibit contrast difference detection). The contrast ratio (CR) is a more direct measure of contrast but does not take into account noise present within the scan. Since noise characteristics can be improved by changes in hardware or by performing multiple signal averages, both metrics allow us to not only compare the scans as they are but also how noise reduction may improve detecting contrast differences. Results Contrast Study Qualitative Analysis The T1-w MP-RAGE subcortical image rev ealed relatively poor contrast among the targets for DBS, however the sequence did allow reasonable localization of striatum and thalamus. T2-w FLAIR scans demonstrated better contrast and were better able to localize the STN, SNr, red nucleus (RN) and pallidum (GPe/GPi). The FGATIR scans allowed for localization of the thalamus, striatum, GPe/GP i, RN, and SNr and displayed sharper delineation of these structures (Figure 3-1). Th e FGATIR revealed features not 43

PAGE 44

visible on other scan types: the internal lamina of the GPi (Figure 3-4C, arrow), fiber bundles from the internal capsule piercing the striatum, and the boundaries of the STN. The T2-w FLAIR sequences poorly imaged t he lateral border of STN. Both the FGATIR and T2-w FLAIR (Figure 3-2C) displa yed the STN as a hypointense structure. However, SNr was hyperintense in the FGATIR which created a degree of STN/SNr contrast not seen in the T2-w FLAIR. SNr was better visualized by FGATIR than by T2-w FLAIR, and was shown as a hyperinte nse region inferior to the STN (Figure 3-2C,D). On FGATIR imaging, the lateral and posterior boundaries of STN were also more distinctly hypointense as compared to their appearance on T2-w FLAIR (Figure 3-2C,D) which appeared to underpredict the boundaries of STN (according to atlas boundaries). While the lateral boundary is mo re consistently hypointense on the FGATIR, the definition between the STN and the PLIC was more difficult to visualize than in the T2-w FLAIR. Intraoperative MER mapping confirmed the region of hypointensity that corresponded elec trophysiologically to STN. The FGATIR also produced higher contrast along the lateral boundary of the thalamus than was seen on T2-w FLAIR or standard T1-w scans. It may also be significant that the VL thalamus can be parti ally distinguished from other thalamic nuclei on FGATIR scans (Figure 3-3C). MER data was consistent with the thalamic boundaries seen with FGATIR both spatially and with respect to sensorimotor somatotopy. The pallidum was hyperintense relative to t he posterior limb of the internal capsule (PLIC) on FGATIR imaging. Th ough both the GPi and striatum were hyperintense they were readily distinguishable (F igure 3-4C) due to their diffe ring average intensities. The boundaries of the GPi and GPe fit with pr edicted atlas boundaries and MER data. 44

PAGE 45

Standard T1 images (Figure 3-4A) showed little contrast between the GPi, GPe, or the PLIC, but there was visible differentiation of pallidum and striatum with this sequence. T2-w FLAIR performed better than standard T1 (Figure 3-4B), showing hypointensity around pallidum relative to striatum and the PLIC, but had less correspondence with atlas data than was seen with FGATIR. T2-w FLAIR failed to distinguish GPi and GPe and seem to overpredict the extent of pallidum medially and laterally (versus atlas predicted boundaries). Quantitative Analysis Figure 3-5 summarizes the CRs and CNRs calculated for these three pilot MRI scans with respect to the DBS targets. In all cases the CNR and CR were higher when using the FGATIR scan than with either T1-w or T2-w FLAIR imaging, with the FGATIR showing the highest level of cont rast across the regions analyzed. Discussion and Conclusi ons Contrast Study The DBS surgical procedure is focused primarily on obtaining accurate electrode placement. Imaging has played a central role in making this a reality. Progress in imagingthe stereotome, CT scans, and MRIhave all advanced DBS by making target localization more precise. Both T1 and T2 weighted scans have been employed in this regard. T1 weighted imaging has been a widely used scann ing procedure for stereotactic surgery and radiosurgery due to its ability to produce thin slice, high resolution acquisitions within relatively s hort time periods. In our experience, T2 weighted MRI has been the primary high-cont rast imaging modality employed in DBS targeting for other groups due to its ability to visualize subcortical structures with high iron concentration. However the resolution and slice thickness have proven suboptimal. The FGATIR sequence in this pilot study offers significant advantages over both 45

PAGE 46

standard T1-w and T2-w FLAIR imaging for the three main targets utilized for DBS in movement disorders. Previous efforts by others to image the basal ganglia by T2 weighted imaging have yielded mixed results. Good in-plane vi sualization of the STN with T2-w scan sequences has been reported by multiple groups (Dormont et al., 2004; Kitajima et al., 2008; Slavin et al., 2006). Slavin and colleagues were able to identify the hypointense STN at 3T on relatively thin (1.5 mm) slic e T2-w images. This required multiple-planes of acquisitions for high resolu tion localization with a scanning time of approximately 30 minutes, as compared to just over eleven minutes with the FGATIR. Other studies have tackled the problem of T2 localization of t he STN, and concluded that the hypointensity usually taken to represent STN sometimes represents solely the medial portion of the STN (Dormont et al., 2004). T2 visualizati on of the STN is also made problematic by interpatient variation in iron deposition in the basal ganglia t hat causes the T2 contrast to be visible to the human eye (Dormont et al., 2004). For these reasons, T2-weighted imaging alone does not provide suffi cient or consistent contrast of certain key structures of the basal ganglia. This lack has been com pounded through the thick slice acquisition methods used in 2D FSE, with slices of 2 mm or more being typical. T1 weighted imaging has also proved awkw ard at visualizing structures at the brains center. Standard T1-weighted imaging has been used to acquire thin slice images of 1 mm to 2 mm (using MP-RAGE or turbo field echo sequences) although with seemingly less contrast than with thicke r-slice T2-weighted images. Recent results indicate that T1-weighting itself should provide significant contrast compared to T2-weighted imaging for the thalamus, thalam ic subnuclei (Deoni et al., 2005a; Mercado 46

PAGE 47

et al., 2006), and the STN (Deoni et al., 2005a). Deoni and colleagues acquired an ultra-high resolution T1 map that was able to delineate thalamic regions but at acquisition times that were fa r too long (several hours) fo r easy clinical application (Deoni et al., 2005a). In a paper by Reich et al. (Reich et al., 2000), the authors described a 2D fast spin echo inversion re covery (FSE-IR) sequence. While they were able to identify the GPi, the resolution wa s low and the slice thickness was relatively thick (2 mm) due to the use of a 2D FSE based inversion recovery sequence. This sequence required 27.5 minutes of scanning time in order to allow for multiple planes of acquisition and sufficient averaging as opposed to the typical whole brain acquisition of an MP-RAGE sequence which is typically in the range of 5 to 15 minutes. The work of these groups shows the potential for T1 as a vehicle for im aging the basal ganglia, but a high resolution, high contrast, thin slice scan in clinically feasible scan time has yet to be demonstrated. Our work focused on optimizat ion of the T1 contrast mechanism. The FGATIR was based on a standard T1-w MP-RAGE sequence but incorporated the idea of nullification found in FLAIR and short tau inversion recover (STIR): nullification of CSF signal (as in FLAIR) and nullifying the signal from fat (as in STIR). This allowed white-matter signal nullification and an image that, in brain parenc hyma, resulted from a gray matter only signal. Though the concept of nullification is not completely novel, the FGATIRs optimization of the full invers ion pre-pulse prior (to nullify the white matter signal) to a fast 3D acquisition sequence may produce fast high-resolution and thin slice scans as compared to either standard 2D FSE or 2D FSE-IR based sequences. 47

PAGE 48

The FGATIRs preferential nullif ication of white matter si gnal allows for excellent delineation of grey matter st ructures that are surrounded by highly myelinated areas, such as structures in the basal ganglia. Fo r example, the thalamus is bounded laterally by the internal capsule, superiorly by t he corona radiata, and inferiorly by the zona incerta and afferent brainstem and cerebellar tracts. The GPi is a grey matter structure bordered on its posterior and medial extent by the internal capsule and anteriorly by the heavily myelinated lamina between GPi and GPe. As can be seen from Figures 3-1 to 3-5, the contrast ratio for t he FGATIR sequence is in general higher than that seen by the other two acquired sequences. Though the STN is well visualized on the FG ATIR, the reasons for its visualization are less clear. The STN is a gray matter structure and appears hypointense on our scan sequence. This may represent a high degree of myelination relative to neighboring grey matter structures. Similar contrast has been r eported with STIR images (Kitajima et al., 2008). It is interesting to note that the VL thalamus is similarly darker than its surroundings on FGATIR. This may similarly suggest that the VL thalamus has more white matter connections than the non-VL thalamus. The quantitative results from this pilot work are in line with our qualitative results (Figure 3-5). The most dramatic CNR was found between STN and t he thalamus on the FGATIR sequence (17.02. versus 7.85 on T2 FLAIR). While T2 images are currently considered to be the best clinical scan for th e localization of STN by most other groups, the CNR between the STN and SNr was 1.76 for the T2-weighted FLAIR scan versus 6.06 on the FGATIR scan. The distinct boundary between STN and SNr has been the most useful boundary in localization of STN on the FGATIR. 48

PAGE 49

It should be noted that since FLAIR and FG ATIR contrast between STN and SNr is different, the STN may at first appear th inner on the FGATIR than on the FLAIR. The size difference may be more apparent t han actual. Both the STN and SNr are hypointense on FLAIR images, creating an area of low contrast between STN and SNr (as seen in contrast measuremen ts in Figure 3-5), which can ma ke it difficult to see the STNs inferior border. Referring to the sagitt al slice (Figure 3-2D) helps rectify this, revealing a hypointense region superior to SNr and inferior to ZI th at corresponds well to the atlas predicted site and size of ST N. The atlas also suggests better FGATIR correspondence to the STN outline than seen on T2 FLAIR. On FLAIR images, the STN (hypointense region within the red contour seen in Figure 3-2B on the left hemisphere) is smaller than the atlas predicted region an d only fills a small portion of the atlas contour for STN. On FGATIR images, STN (h ypointense) seem to be of at least the same size as compared to T2 imaging and al so appears to better ma tch the size of the atlas predicted region in the lateral extent. Figure 3-2B and 3-2C show a red line (of equivalent length in both panels) on the ri ght hemisphere (left side of image) that represents the superior to in ferior extents of the STN for both FLAIR and FGATIR imaging. The hypointense region in both the FGATIR and FLAIR appear to be roughly the same size in the superior to inferior boundary. While the boundary between ZI and STN is an important one to identify clearly, as it marks the STNs superio r boundary, there can be seen heterogeneous intensities in the superior STN and inferior ZI region. The origin of thes e is unclear. From Figure 3-2C, it can be seen that ther e is a thicker region of hypoi ntensity just inferior to thalamus and superior to the ST N which, on the atlas, maps onto ZI. This hypointensity 49

PAGE 50

fits well with the description of ZI as an ax on-rich structure. Based on empirical data from T1 images and our MER mapping data, the hypointense region directly superior to SNr appears to map to STN. The origin of the hyperintense r egion found immediately superior to STN is unclear. Our atlas deforma tion technique consistently maps it to superior STN/inferior ZI, but the appropriate interpretation of the heterogeneity of the MR signal in this area is currently indete rminate. Evaluation of the hyperintensity between ZI and STN in the FGATIR may be po ssible via an imaging/ histological study of a fresh cadaver brain. DBS imaging has generally fallen into two seemingly mutually exclusive categories: high resolution thin slice MRI and high contrast thick slice MRI. Many groups have focused on obtaining excell ent contrast in plane at t he expense of localization along the slice axis (Dormont et al., 2004; Ki tajima et al., 2008; Reich et al., 2000; Slavin et al., 2006). Figure 3-6 shows a refo rmatting of the FGATIR scan at various axial slice distances: 1 mm (Figure 3-6A,B ), 2 mm (Figure 3-6C ), and 3 mm (Figure 3-6D). The reconstructed views (coronal in Figure 3-6) show worsening resolution as the simulated slice distance is increased. It is worth noting that despite the high contrast seen with 3 mm slices, they are too thick to be useful for direct targeting of STN or GPi as the boundaries in the reconstructed image can no longer be easily delineated (Figure 3-6D). The ability of the FGATIR sequence to acquire thin slice high resolution (less than or equal to 1 mm in all dimensions) imag es is crucial to its utility as a tool for imaging the basal ganglia. The ability of the FGATIR sequence to acquire both thin and high contrast slices is a significant improvement over the current sequences of choice: t he 2D FSE-IR or 2D 50

PAGE 51

FSE T2-w. Using these 2D sequences can limi t slice thickness to a minimum of 2 mm due to hardware gradient specifications. With the FGATIR sequence the slice thickness is no longer hardware gradient li mited. The scan is fully 3D and slice thickness is limited only by acquisition time and signal consider ations. Another advantage of the FGATIR is that at 1 mm slice thickness it offers nearly isotropic voxels. This a llows for multi-planar reconstruction (MPR) and true 3D manipulati on. The anisotropic voxels from 2 mm (or thicker) slice thickness 2D scans offer poor resolution when subjected to these reconstructions (Figure 3-6) and transformations. Our pilot work demonstrates that the contrast between target and surrounding tissue on the FGATIR is superior to st andard T1-w and T2-w FLAIR imaging. This combination of short scan time, high resoluti on, thin slices, and high contrast has not previously been reported. Futu re work will involve the anal ysis of the FGATIR across a larger number of patients to determine the level of detectability across our patient populations. Due to the flexib ility of this scan technique, we may be able to take advantage of newer hardware such as multic hannel phased array head coils in order to increase the SNR as well as implement hig her resolution and even faster acquisitions (through parallel imaging). Our initial results on a Philips 3T Achieva MRI with SENSE indicate that sub-millimeter (0.75 mm) slice thicknesses may be acquired within 10 minutes using an 8-channel headcoil. It is also possible to implement this sequence on standard 1.5T machines, but it should be not ed that (as with any other sequence) the parameters would have to be changed accordingly to produce similar contrast characteristicsthe parameters listed in Table 3-1 have been optimized for 3T implementation. 51

PAGE 52

The consequences of the FGATIR sequence for DBS may prove substantial. Since only a few patient scans have t hus far been evaluated we were unable to quantitatively determine whether the use of th is imaging protocol reduces the number of MER passes. It is our impression that our operating room team was better able to recognize structures during preoperative targeting, increasing the ease of atlas deformation and ease of selection of the coordi nates of the initial MER pass. We expect that the end result of this improved imaging method will be a reduction in operating room time via a reduction of MER passes, intraoperative testing, and potentially DBS lead relocation due to suboptimal initial implantation. Length of time spent in surgery is a well known correlate to surgical complicat ion rates and reducing the time the patient spends on the OR table would be expected to significantly decrease morbidity. We do expect that better direct imaging will decrease the amount of MER required to arrive at target, and wil l examine this question more closely. We also expect improved imaging of the basal ganglia to allow us to better define the degree of interpatient variability in basal ganglia ex tent and structure. Most atlases are based on the careful dissection of a small number of brains. The FGATIR may allow us to substantially expand this data set, improvi ng our ability to understand and adapt to the anatomical variations underlying individual patients disease. Methods Targeting Study Subjects For this study, we used the preoperative scans of twenty patients who were to undergo STN DBS. These patients underwent forma l evaluations by a fellowship-trained Movement Disorders Neurologist, a Neurosurgeon, a Psychiatrist, and a Neuropsychologist in order to ensure accuracy of diagnosis by clinical criteria (Okun et 52

PAGE 53

al., 2004b) as well as absence of significant cogni tive or psychiatric comorbidity. Prior to implantation, these PD pati ents were also required to demonstrate at least a 30% improvement in the motor subsection of the Unified Parkinsons Disease Rating Scale (UPDRS III) between the on and off medi cation states (Okun et al., 2004b). Preoperative Imaging Each included patient received three scans on the day prior to surgery: a three plane localizing scout, a T1-weighted 3D Magnetization PreparedRapid Acquisition Gradient Echo (MP-RAGE), and a T1-weighted 3D FGATIR (the last three scans were each single volume whole brain scans). All scans were acquired on a clinical Siemens Allegra 3T MRI using a quadrature birdcage headcoil. The total scanning time for all three scans was less than 20 minutes. Specific parameters utilized are listed in Table 3-1. Surgical Procedure On the morning of the operation, a Co smanRobertsWells (CRW) head ring was applied under local anesthes ia and a high-resolution stereotactic head computed tomography (CT) scan was performed. The CT and MRI images were fused using in-house computer software, a software package that is analogous to the Varian (Palo Alto, CA) system but with several added features that facilitated ta rgeting in atlas space by coregistering anatomical l andmarks. A Cartesian coordinate system confirmed the patients mid-commissural point and this point was used as a reference to confirm the target (Okun et al., 2007). The software utilized allowed the display and registration of a deformable 3D atlas, based on the Schaltenbrand and Bailey stereot actic atlas (Schaltenbrand and Bailey, 1959) which was then overlaid onto the MRI scans. The atlas and scans were used 53

PAGE 54

together to target the struct ure of interest. We employed a two step targeting process. An initial estimate based on anatomic (AC/PC/midline point) coordinates was then followed by direct targeting adjustments us ing a deformed atlas overlay over image slices as well as direct visualization of ta rget structures. Target selection for all the patients in this study was the STN. These pat ients had their initial target points within the motor (dorsolateral) STN. The indirect at las coordinate for the STN target (tip of lead) for an initial MER pass was approximately AP -3 mm, LT 11 mm, and AX -7 mm. Coordinates were modified from this indirect targeting by using atlas deformation and direct visualization of target structures. The final target was determined through the use of electrophysiological data (MER maps) with refe rence to the initial target estimate. We implanted 3 to 3.5 mm posteri or to the anterior border and 2.5 to 3 mm medial to the internal capsule boundary with the deepest electrode contact placed at the STN/SNr boundary. Targeting Study Procedure In order to simulate the procedure of ta rgeting using various techniques without implanting patients using t hese various techniques, the scans of twenty DBS STN patients were used for the sole purpose of targeting under various methods. Two advanced neurosurgery residents (one with a year of DBS fe llowship training and one without specific DBS training) were ask ed to target these twenty patients without knowledge of where the electr ode was clinically implanted in these patients. They targeted using four different methods (in order): 1) CT only, 2) CT + MP-RAGE, 3) CT + MP-RAGE + atlas, and 4) CT + MP-RAG E + atlas + FGATIR. The neurosurgery resident who had no previous DBS training was asked to target a second time using the CT only method to assess the effe ct of learning on the process. 54

PAGE 55

This stepwise targeting procedure was chosen to emulate the chronological advancements in targeting for DBS as practiced at STH. For all methods, the neurosurgery residents were told to target the position of final implantation which was (as discussed previously) chosen to be 3 to 3.5 mm posterior to the anterior border and 2.5 to 3 mm medial to the internal capsule boundary with the deepest electrode contact placed at the STN/SNr boundary. Targeting was accomplished by first setting up an anatomic coordinate system based off the AC, PC and a midline point. These points were chosen off the CT scan in the case of the first methods and the T1w MP-RAGE in all the subsequent methods. Subsequently, a target point was picked that placed the tip of the electrode at t he position previously mentioned and at a reasonable angle that did not pass through the ventricles, sulci, or bl ood vessels for safety reasons. The optimal angle was chosen considering these safety issues while still placing the third contact of the electrode just inferior to the boundary be tween STN and ZI. In methods 3 and 4, the in-house created 3D atlas (previously discussed) was deformed (by the neurosurgery resident) to fit the patients anatomy as well as possible for the imaging methods available (T1w MP-RAGE and/or FGATIR). In all cases, the residents were instructed to start with an initial target based off the anat omic coordinate system and to refine that target as best as possible using available imaging (or atlas) methods. If the residents were able to directly or indirectly vis ualize the boundaries of ST N, they could refine placement of the target positi on to achieve placement at 3 to 3.5 mm posterior to the anterior border and 2.5 to 3 mm medial to the internal capsule boundary with the deepest electrode contact placed at the STN/ SNr boundary. If there was insufficient imaging contrast to adjust fo r patient anatomy, the placement would be primarily based 55

PAGE 56

off the anatomic coordinate system All twenty subjects were targeted first using method 1 then followed by 2 and so on and so forth with time (days or weeks) between each step to reduce the effect of l earning and memorization on targeting. Analysis of Target Points After all the targets had been chosen by t he residents, targeti ng data from each method was tabulated. These data were co mpared with the clinical final target placement data for each of these particular patients. The clinical final target was determined through a combinat ion of initial targeting with imaging data augmented with knowledge of the electrophysi ology of the patient. The electrophysiology (MER data) map of each patient was used to directly confirm boundaries of the STN. Differences between the clinical final target location an d targets picked by each of the neurosurgery residents (for each method) we re calculated. These differences were calculated as absolute errors and signed errors for the anterior to posterior, medial to lateral, and superior to inferior directions. The signed errors were used to determine if any bias/systematic error ex isted in the dataset. Results Targeting Study Targeting data from the twenty STN DBS patients were tabulated and the clinical final target was used as the basis for co mparison against the experimentally chosen targets of the two neurosurgery residents. The four methods (discussed previously) were each independently compared against the clinical final target for each patient. Figure 3-7 shows the magnitude difference (error) between the target point picked by the resident and the clinical target point al ong the three axes ( AP, LT, and AX). The mean errors for the DBS trained resident r anged from about half a millimeter to about one millimeter (Figure 3-7A). The mean errors for the re sident without any prior training 56

PAGE 57

tended to be higher ranging between about a half millimeter and 1.5 mm with one exception (Figure 3-7B). The AP mean error associated with targeting by the fourth method (CT+MP-RAGE+Atlas+FGATIR) had an error of 2.63 mm (Figure 3-7B). Examination of the bias in targeting of that particular method revealed a large systematic bias and this was further looked at in Figure 3-8. The standard deviation of the errors is shown in both panels as the error bars. The potential for systematic targeting error in the data was examined by subtracting out the average (sig ned) error. If bias was present in the data, this method should reduce it. In general, the magnitude errors decreased following removal of systematic bias (Figure 3-8). In both ca ses (with and without removal of systematic error), the DBS trained resident tended to have improved targeting with the fourth method (CT+MP-RAGE+Atlas+FGATIR) versus the first method (CT only) but this improvement was not stat istically significant in either ca se (Figure 3-8A). The unusually large error in the AP direction using the fourth method of tar geting by the second resident (seen in Figure 3-7B) was reduced to within 1.5 mm once systematic bias was accounted for (Figure 3-8B). Comparisons of targeting using method 1 versus any other method does not appear to be significantly better especially considering the large standard deviations in the errors (error bars seen in Figure 3-7 and 3-8). There does not appear to be any statistically significant im provement on targeting through the addition of imaging or atlas methods. Discussion and Conclusi ons Targeting Study The purpose of the targeting study was to evaluate utility of various targeting methods as they would be used clinically at UF and Shands. In order to accomplish this goal, the examined methods were chosen as anal ogs to previous targeting techniques 57

PAGE 58

used at Shands. The final target met hod (CT+MP-RAGE+Atlas+FGATIR) examined was chosen as a method to validate the clin ical utility of the FGATIR sequence for the STN. The previous sections (regarding the contrast study) of th is chapter examined specifically contrast differences across diffe rent targets for DBS (including the STN). This targeting study was an attempt to eval uate the overall effect of imaging methods on targeting. From Figures 3-7 and 3-8, there does not appear to be any significant advantage to using any of the more advanced techniques over CT for targeting of the STN. This outcome is in disparity to the results prev iously discussed in the contrast study where the FGATIR sequence provided significantly higher contrast of the STN compared with other imaging methods. The reason for this result may lie in a combination of several reasons. One possibility is that the STN does not vary significantly in size, shape, or location relative to the two anatomic landmarks used in this study (AC and PC). As discussed in chapter two, the STN is a small s ubcortical structure that is relatively close to midline and lies between the two landmarks in the AP direction used to establish the anatomic coordinate system. It has not been well established how much anatomic variation there is in the size, shape, and locati on of this structure especially relative to diseased patients where disease progression may affect these parameters. In the case of the CT only method, the neurosurgery residents often picke d the target based solely off AC and PC coordinates as there was little to no contrast in the imaging to use for direct visualization/targeting. Since more di rect (image based) targeting did not perform significantly better or worse than less direct (coordinate based) targeting methods, the deviation of the STN from ideal coordinat es may not be significant. This issue of 58

PAGE 59

localization accuracy of indire ct targeting through landmarks versus direct localization of the STN has been previously examined by Cuny et al (Cuny et al., 2002). In their paper, localization of the STN was accomplished through direct (T2 weighted imaging) and indirect (AC/PC coordinate based) methods. The most accurate targeting method was found to be indirect localization using AC/PC coordinates although the resolution of the different methods used were not comparable, the direct method resolution ( 2 mm slice thickness) was poor compared to the indi rect localization method (high resolution ventriculography or high resolution thin slic e T1 weighted imaging). In our work, the deviations of the resident picked targets versus clinically picked final targets are similar to the errors seen between clinically planned final targets and the actual post-operative measured electrode positions within the patient s brain (AP: 1.12 1.03 mm, LT: 1.42 1.02 mm, AX: 0.85 0.65 mm). Since these mean errors were generally around 1 mm in each direction, these errors are roughly wi thin the error associated with being able to physically place an electrode within a patients brain for this target. While more advanced methods for target ing were not found to conclusively provide any tangible benefit in the targeting of STN, this result may be partially the result of the target chosen. The STN target was the most often targeted structure at Shands. Future studies of other targets may stand to benefit from using atlas methods and/or FGATIR imaging. From the previously discussed contrast study, there was significantly higher contrast of subcortical structures in general within the FGATIR images compared with other imaging methods. Additional targeting studies of GPi, thalamus, or even other areas of the brain may reveal the utility of these more advanced targeting methods (atlas, FGATIR, etc) versus indirect targeting. 59

PAGE 60

Table 3-1. Parameters used for the scans in this study. T1-w 3D MP-RAGE T2-w 3D FLAIR T1-w 3D FGATIR Repetition time (TR) 1600 ms 6000 ms 3000 ms Echo time (TE) 4.38 ms 353 ms 4.39 ms Inversion time (TI) 800 ms 2200 ms 409 ms Inversion pulse angle 90 180 180 Matrix 384x288 256x240 320x256 Field of View (mm) 256x192 256x240 256x192 Slices 160x1 mm 160x1 mm 160x1 mm Orientation Axial Sagittal Axial Bandwidth 130 Hz/Px 1302 Hz/Px 130 Hz/Px Acquisition time 6:45 minutes 12:08 minutes 11:14 minutes 60

PAGE 61

Figure 3-1. Sagittal images of subcortical stru ctures in the A) T1-w 3D MP-RAGE, B) T2-w 3D FLAIR, C) T1-w FGATIR, and D) T1-w FGATIR with deformable atlas contours overlaid. The contour colo rs for the deformable atlas are (from most anterior to most posterior): st riatum (blue), GPe (green), anterior commissure (black), GPi (red), optic trac t (yellow), thalamus (green), various VL thalamic nuclei (green), STN (red), and SNr (black). 61

PAGE 62

Figure 3-2. Subcortical images from a STN DBS patient (pre-surgery) showing coronal slices with deformable atlas contour ov erlay on the A) T1-w 3D MP-RAGE, B) T2-w 3D FLAIR, C) T1-w FGATIR, and a sagittal slice through the D) T1-w FGATIR with MER maps overlaid (r epresented by colored dots). The white arrows represent the location of STN in the coronal and sagittal slices. The red lines in panels B and C (both are the same length) represent the distance (superior to inferior) of the STN in both FLAIR and FGATIR images. Although STN may appear thinner in the FGATIR, as can be seen from the line it is roughly the same size in FLAIR and FGATIR images. In panel D, the dot colors represent cells found for STN (red), thalamus (green), ZI (black and labeled as ZI), and SNr (black and unlabeled). Squares represent regions at which cellular response was seen during either passive movement or sensory stimulation of the patient. 62

PAGE 63

Figure 3-3. Subcortical images from a thalam ic Vim DBS patient (pre-surgery) showing axial slices (with deformabl e atlas overlay) on the A) T1-w 3D MP-RAGE, B) T2-w 3D FLAIR, C) T1-w FGATIR, and a sagittal slice through the D) T1-w FGATIR with MER maps overlaid (r epresented by the colored dots). The green contours within the thalamus represent the: (from anterior to poster) ventralis oralis anterior (Voa), ventra lis oralis posterior (Vop), ventralis intermedius (Vim), and ventralis caudalis (Vc) nuclei of the thalamus. In panel d, the dot colors represent cells found for thalamus (green). Squares represent regions at which cellular response was seen during either passive movement or sensory stim ulation of the patient. 63

PAGE 64

Figure 3-4. Subcortical images from a GPi DBS patient (pre-surgery) showing axial slices (with deformable atlas overlay) on the A) T1-w 3D MP-RAGE, B) T2-w 3D FLAIR, C) T1-w FGATIR, and a sagitta l slice through the D) T1-w FGATIR with MER maps overlaid (represented by the colored dots). The white arrows represent the location of the lamina between GPi/GPe and the lamina within GPi. In panel D, the dot colors re present cells found in GPi (red), GPe (green), striatum (blue), and optic tract (yellow). 64

PAGE 65

Figure 3-5. Contrast rati os (CR) for GPi/PLIC, STN/Thalamus, STN/SNr, VL Thalamus/non-VL Thalamus, and VL Thal amus/PLIC. The contrast ratio of the FGATIR sequence is higher than seen with either the T1-w or the T2-w FLAIR scans. Numerical values for both CR and CNR are listed. 65

PAGE 66

Figure 3-6. Simulated FGATIR images at A,B) 1 mm, C) 2 mm, and D) 3 mm slice spacing. The superior and inferior boundaries of STN are visible in A) and B) as shown by the white arrows in panel A, but are difficult to identify in C) and D). 66

PAGE 67

A 0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60 1.80 2.00Error (mm) CT Only CT+MP-RAGE CT+MP-RAGE+Atlas CT+MP-RAGE+Atlas+FGATIR CT Only 1.05 1.13 1.07 CT+MP-RAGE 1.09 1.03 0.63 CT+MP-RAGE+Atlas 1.10 1.01 0.79 CT+MP-RAGE+Atlas+FGATIR 0.94 0.94 0.53 AP LT AX B 0.00 0.50 1.00 1.50 2.00 2.50 3.00Error (mm) CT Only (First) CT+MP-RAGE CT+MP-RAGE+Atlas CT+MP-RAGE+Atlas+FGATIR CT Only (Second) CT Only (First) 1.24 1.01 1.40 CT+MP-RAGE 1.02 1.09 0.57 CT+MP-RAGE+Atlas 1.06 0.99 0.66 CT+MP-RAGE+Atlas+FGATIR 2.63 0.86 1.03 CT Only (Second) 1.12 1.09 1.57 AP LT AX Figure 3-7. Absolute errors in target planni ng using various targeti ng methods for A) the DBS trained neurosurgery resident and B) the neurosurgery resident without previous training. Errors are one standard deviation. 67

PAGE 68

A 0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60 1.80 2.00Error (mm) CT Only CT+MP-RAGE CT+MP-RAGE+Atlas CT+MP-RAGE+Atlas+FGATIR CT Only 1.05 0.89 0.69 CT+MP-RAGE 1.08 0.84 0.64 CT+MP-RAGE+Atlas 1.07 0.85 0.62 CT+MP-RAGE+Atlas+FGATIR 0.94 0.70 0.42 AP LT AX B 0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60 1.80 2.00Error (mm) CT Only (First) CT+MP-RAGE CT+MP-RAGE+Atlas CT+MP-RAGE+Atlas+FGATIR CT Only (Second) CT Only (First) 1.26 0.83 0.98 CT+MP-RAGE 1.02 0.99 0.53 CT+MP-RAGE+Atlas 1.06 0.90 0.65 CT+MP-RAGE+Atlas+FGATIR 1.30 0.75 0.80 CT Only (Second) 1.13 1.01 0.77 AP LT AX Figure 3-8. Corrected (for bias) absolute errors in target planning us ing various targeting methods for A) the DBS trained neurosurgery resident and B) the neurosurgery resident wit hout previous training. Errors are one standard deviation. 68

PAGE 69

CHAPTER 4 LOCALIZATION OF FUNCTIONAL TERRIT ORIES OF SUBCORTICAL STRUCTURES Aims Our aim was to compare probabilistic diffu sion tractography (PDT) derived subregions within the thalamus with microelectrode data and atlas regions. Background Many of the structures targeted in DBS have multiple distinct segregated circuits that comprise it (Alexander et al., 1990; Alexander et al., 1986). Diffusion weighted imaging is a type of scan that can be used to potentially image and follow these tracts/circuits in a non-invasive manner. Diffusion weighted imaging as a whole has been a significant development to the field of medicine. Some of the earliest clinically relevant work in diffusion imaging used diffu sion weighted imaging to detect ischemic tissue (Moseley et al., 1990b). Diffusion we ighted imaging has also been used to quantify the isotropic/anisotropi c nature of water diffusion in vivo in a cat spinal cord (Moseley et al., 1990a) and in cerebral whit e matter (Chenevert et al., 1990; Doran et al., 1990). Characterization of water diffusion anisotropy has since been modeled using various methods. The most commonly used model was developed by Basser (Basser et al., 1994) in which a tensor was used to model Gaussian diffusion of anisotropy. In this particular model, the primary eigenvector is typically taken as the dominant fiber orientation and used as a basis for streamline tractography algorithms. In order to fit a rank-2 tensor to each voxel, a measurement of six distinct diffusion directions is necessary. Diffusion measurement along a much higher number of directions has since been proposed as a way to increase the model complexity of diffusion. Other models can be assumed using high angular resoluti on diffusion imaging (HARDI) (Tuch et al., 69

PAGE 70

2002) acquisition methods. Diffusion models such as Q-ball (Tuch, 2004) and probabilistic diffusion (Behrens et al., 2003b) as well as numerous others have since been used as alternative models to the standard diffusion tensor model in the attempt to more fully model the diffusion anisotropy at each voxel. Using diffusion models such as diffusion tensor imaging (DTI) and associated tractography, visualization and localization of these tracts has been used examined by multiple previous groups as a way to locali ze targets in DBS (Aravamuthan et al., 2008; Gutman et al., 2009; Owen et al., 2007; Sedrak et al., 2008). Since the thalamus has rich connections to the cortex, it has specifically been looked at by several groups using diffusion imaging and tractography to try and separate out i ndividual thalamic nuclei (Behrens et al., 2003a; Wiegell et al., 2003; Yamada et al., 2009). These groups have parcellated regions of the thalamus through either DTI pr imary eigenvector (Wiegell et al., 2003) or tractography of connections to the cortex (Behrens et al., 2003a; Yamada et al., 2009) and found that they generally relate to large thalamic nuclei. In this work, the particular small thalamic nuclei of intere st for DBS in essential tremor (Vc and Vim nucleus) were localized using previously developed probabilistic diffusion tractography (PDT) methods (Behrens et al., 2003b). We co mpared this method against clinical data (deformed atlas structures and microelectrode data) in five essential tremor patients. Materials and Methods Subjects For this study we scanned five patients with essential tremor who were to undergo DBS. These patients underwent formal evaluat ions by a fellowship-trained Movement Disorders Neurologist, a Neurosurgeon, a Psychiatrist, and a Neuropsychologist in order to ensure accuracy of diagnosis by clin ical criteria (Okun et al., 2004b) as well as 70

PAGE 71

absence of significant cognitive or psychiatric comorbidity. All patients were implanted with DBS electrodes on their left si de (all patients were right handed). Preoperative Imaging Each included patient received four scans on the day prior to surgery: a three plane localizing scout, a T1-weighted 3D Magnetization PreparedRapid Acquisition Gradient Echo (MP-RAGE), a T1-weighted 3D Fast Gray Matter Acquisition T1 Inversion Recovery (FGATIR) (these two scans were each single volume whole brain scans), and a 12-direction diffusion tensor imaging (DTI) protocol. All scans were acquired on a clinical Siemens Allegra 3T MRI using a quadrature birdcage headcoil. The total scanning time for all four scans was approximately 35 minutes. Diffusion weighted data were acquired using a spin-echo echo planar imaging sequence (SE-EPI) (70 1.6 mm th ick axial slices, matrix si ze 128 128, field of view 205 205 mm 2 giving an isotropic voxel size of 1.6 1.6 1.6 mm 3 ). The diffusion weighting was acquired isotropically along 12 separate directions with a bvalue of 1000 s mm -2 Twelve diffusion directions were acquired as this was the highest number of diffusion directions acquirable on this version of the clinical Siemens 3T Allegra scanner software used for all these scans. For both the diffusion weighted and zero weighted data, the repetition time (TR) was 8600 ms wit h an echo time (TE) of 76 ms. Five sets (one zero diffusion weighted image and twelve diffusion weighted images in each set) were acquired with two averages for each set in order to attain adequate signal-to-noise ratio (SNR). Surgical Procedure On the morning of the operation, a Co smanRobertsWells (CRW) head ring was applied under local anesthes ia and a high-resolution stereotactic head computed 71

PAGE 72

tomography (CT) scan was performed. The CT and MRI images were fused using in-house computer software, a software pa ckage that is analogous to the Varian FastPlan (Palo Alto, CA) system but with se veral added features, that facilitated targeting in atlas space by coregist ering anatomical l andmarks. A Cartesian coordinate system confirmed t he patients mid-commissural point and this point was used as a reference to confirm the target (Okun et al., 2007). The software utilized allowed the display and registration of a deformable 3D atlas, based on the Schaltenbrand and Bailey stereot actic atlas (Schaltenbrand and Bailey, 1959) which was then overlaid onto the MRI scans. The atlas and scans were used together to target the struct ure of interest. We employed a two step targeting process. An initial estimate based on anatomic (AC/PC/midline point) coordinates was then followed by direct targeting adjustments us ing a deformed atlas overlay over image slices as well as direct visualization of ta rget structures. The patients in this study had the anterior boundary of the ventra lis caudalis nucleus (Vc) of the thalamus targeted. This boundary was estimated based on indirect targeting and confirmed by the use of microelectrode recording (MER, detailed belo w). The final electrode location was 2 mm anterior to this point along t he anterior boundary of the ventral intermediate nucleus of the thalamus (Vim) and the ventralis oralis posterior (Vop) nucleus. Coordinates were modified from this indirect targeting by using atlas deformation and direct visualization of target structures. Microelectrode Recording and Registration The target coordinates were verified to be within the region of interest via one or more MER passes. Our technique used a 3-D mapping procedure to guide electrode placement. For each pass, cellular activity was recorded at millimeter intervals 72

PAGE 73

beginning at 30 mm above the selected target, and at submillimeter intervals as the microelectrode approached the ta rget region. At each inte rval the encountered region was determined by the recording neurologist based on the sound and appearance of the recording and the depth at which it wa s observed. Each such determination was represented in real-time as a color-coded point overlaid on an individual patients MRI at the corresponding stereotactic coordinates. In addition to single cell recordings, cellular firing in response to passive motion and sensor y stimulation was used to delineate the somatotopic organization of the target structure. The use of registration software allowe d the translation of MER passes into a linear map of structure and so matotopy that was then over laid on a patients MRI. A final decision, based on this aggregate map, was made as to the optimal location to place the permanent DBS elec trode. The procedure decision -making process took into account both electrode tip location within the target region and the el ectrodes proximity to regions near the target that might result in side effects when stimulated. The electrode is inserted at a slightly sha llower angle than that of the plane of the typical Vc/Vim boundary. This allows us to locate the anterior border of Vc as we pass through the Vim side of the Vc/Vim border superiorly and the Vc side of that border inferiorly. The transition from Vim to Vc can be appreciated as a change from motor-responsive to sensory-responsive cell s. We typically implant the stimulating electrode 2 mm anterior to the hand region of Vc. Diffusion Data Processing In total, 65 volumes of diffusion data were acquired for each patient (5 non-diffusion weighted and 5x12 diffusion weight ed). All of the diffusion data were motion corrected relative to the first non-di ffusion weighted image. The predicted motion 73

PAGE 74

parameters were used to correct the diffusion gradient directions relative to their original imaging plane. No patient had an angular motion greater than 2 degrees from the first non-diffusion weighted image. These data were then used in the FMRIB Software Library (FSL) software package to determine t he parameters for a tw o-fiber field model at each voxel (Behrens et al., 2003b). This particular model was employed in our work for tractography as we acquired greater t han six distinct diffusion directions which allowed for a model more complex than a st andard rank-2 tensor. In addition, Behrens et al has previously used the probabilistic di ffusion model for tractography to parcellate the thalamus according to its cortical connections (Behrens et al., 2003a). Although increasingly higher order models can be used to model the crossing/branching of fibers during tractography, the parcellation of the thalamus into thalamic nuclei does not appear to benefit greatly from a multi-fiber model versus a single fiber model (Behrens et al., 2007). This model was then used with FSL softwar e to perform probabilistic tractography. In the ideal case where the local fiber ori entation has no uncertainty, this probabilistic tractography methodology amounts to a si mple streamline algorithm. For actual diffusion data, samples are taken at each vo xel from the posterior distribution on principal diffusion directions and the streamli ne progresses in that direction. As such, two samples arriving at the same voxel are likely to leave along slightly different directions. This scheme allows for a repres entation of the relative probability of a streamline passing through a voxel. Several regions were manually segmented for usage as masks in the tractography procedure. All diffusion related processing was completed in the original diffusion space 74

PAGE 75

for each of the patients. The cortical ma sks were created by manually contouring the somatosensory cortex (S1) and primary motor cortex (M1) on the non-diffusion weighted image. The anatomic la ndmark used to delineate S1 was the posterior bank of the central sulcus plus the post central gyrus, from the dorsal surfac e of the brain to the lateral fissure. The anatomic landmark used to define M1 was the anterior bank of the central sulcus and posterior half of the prec entral gyrus. These anatomic definitions have been used by previous groups (Behr ens et al., 2003a) for the purpose of examination of thalamic connec tions to cortical areas. An example of these regions of interest can be seen in a near midline sagitta l plane in Figure 4-1. A thalamus mask was created by using the atlas defi ned region described previously. Probabilistic diffusion tractography was comp leted by seeding each voxels in the thalamus mask with 5000 samples with a separ ate analysis for each of the cortical masks (S1 and M1) as a waypoint. For each of these tracts, we looked at the correlation of the DTI derived CDR versus intraoperatively acquired MER data, deformed atlas regions, and clinical stimulation data on an individual basis. Two methods of defining a CDR were examined relative to each other, ei ther by pure thresholding (allowing for interdigitation) or by max fiber association (considering regi ons without interdigitation). The former models areas within the seed r egion (thalamus) as potentially having innervations from multiple cortical regions whereas the latter is a simpler model where voxels are considered to strictly receive innervations from singl e cortical regions. Results Hard versus Soft Delineation of Connectivity Defined Regions (CDR) Previous work (Behrens et al., 2003a) has typically defined CDRs by the origination of the fiber s with the greatest intensity. For ex ample in a particular thalamic 75

PAGE 76

voxel, probabilistic tractography results at a particular voxe l may show connections from both S1 and M1 cortex but that particular voxe l would be classified as belonging to a particular CDR by the particular cortical region which had t he most significant innervation (either S1 or M1 but not bot h) through tractography. In this work, we consider the case of both the standard cl assification of a rigidly defined CDR and an alternative definition for a CDR in which thes e boundaries are not considered to be rigid. In this alternative definition, a particu lar voxel may correspond to multiple CDRs simultaneously depending on the extent of tractography innervation from corresponding cortical regions. We examined th e difference in location of ri gid (hard) versus non-rigid (soft) delineations of CDRs. Microelectrode Data versus Diffusion Tractography Microelectrode data that was acquired intraoperatively was localized to stereotactic coordinates withi n each subject allowing us to determine the location of cells relative to the subjects imaging data. In general, four types of cells were found: non-thalamic cells, thalamic cells with no detectable sensor imotor response, thalamic cells with a passive (proprioceptive) moto r response, and thalamic cells with a sensory response. If the cell had either a sensory or a motor response, the body part location associated with the s ensorimotor responsiv e cell was recorded. For the S1 and M1 tracts through the t halamus created from diffusion data, we examined the fiber intensity at the point for each of the sensorimotor responsive cells. The fiber intensity at a particular voxel is defined as the total number of tracts that pass through that voxel that start from the seed region and go through the particular cortical area of interest (either S1 or M1 ). Four analyses were completed across all 5 subjects and are shown in Figure 4-2 as cumulative histograms depicting: S1 fiber 76

PAGE 77

intensity at sensory responsive cell positions (Figure 4-2A), S1 fiber intensity at passive motor responsive cell positions (Figure 4-2A), M1 fiber intensity at sensory responsive cell positions (Figure 4-2B), and M1 fiber intensity at passive motor responsive cell positions (Figure 4-2B). While the optimal situation would be to have sensory and passive motor responsive cells only residing within S1 and M1 tracts, respectively, the sensory and passive motor responsive cells we found from microelectrode data were not necessarily well spatially separated from each other with sensory and motor cells often being detected at the sa me position. The overall trend of sensory versus passive motor responsive cells suggests that sens ory cells are situated more posteriorly compared to the passive motor responsive cells. From Figure 4-2A, we find that a threshold fiber intensity value of around 800 to 1000 provides good coverage of sensory responsive cells (type I error) with minimal coverage of passive motor responsive cells (type II error). At an S1 fiber intensity thre shold of 800, 88.14% of all sensory cells are within the threshold (out of 59 cells) while in cluding only 33.7% of all passive motor cells (out of 92). At an S1 fiber intensity thres hold of 1000, 83.05% of all sensory cells are within the threshold while including 30.43% of all passive motor cells. More conservative thresholds can be considered to reduce the num ber of motor cells t hat are within the M1 CDR, but at the expense of sensory cell incl usion. In Figure 4-2B, we do not see an analogous threshold of the M1 tract possibl y suggesting that M1 fiber CDRs have strong connections into regions with significant numbers of sensory responsive cells. Significant interdigitation of sensory and passive motor responsive cells was noted at several sites tested intraoperatively with ME R which resulted in regions that had both 77

PAGE 78

sensory and motor responsive cells wit hin the measuring distance of the microelectrode. An exemplar case of tracts from the thalamus to S1 and M1 cortex (with MER overlays) is shown in Figure 4-3A and 4-3B respectively. The images shown are of a para-sagittal FGATIR plane from a single pati ent with one MER pass with thalamic cells noted by the green dots. The cont ours represent atlas structur es such as green for the thalamus and thalamic nuclei. The most poste rior green contour within the thalamus is the Vc nucleus and just anterior to that c ontour is the Vim nucleus (sensory and motor responsive regions, respectively). At a conser vative threshold of 800 (in this particular case), the S1 tract provides significant coverage of sensory responsive cells as noted with t or T for light touch or tactile cells (Figure 4-3A). At the sa me threshold, the M1 tract provided complete cover age of motor responsive cells but also completely covered regions inhabited with sensor y responsive cells. It should be noted that at the same location of sensory responsive cells, passive motor responsive cells were also present in many locations. This occurrence was typi cal across the group of patients. In this particular case, the sensory and passive mo tor responsive cells were not completed contained within the Vc and Vim nuclei, respectively. Rigidly defining the CDR produc es significantly different results in the location and extent of the regions defined for S1 and M1. Si nce there is no analogou s fiber intensity threshold, the voxel is categorized as belonging to a particular CDR by whichever associated fiber is the highest at a particular vo xel. The fiber intensity for M1 tracts was greater than S1 tracts at all MER cell locations (sensory and proprioceptive motor) except for one sensory cell (out of 59 total sensory cells). This rigid definition for 78

PAGE 79

classification correlates poorly with the electrophysiological MER data found from these patients. Atlas Data versus Diffusion Tractography Atlas regions for thalamic nuclei were used as a basis of comparison versus CDRs created from S1 and M1 tracts. Particular atlas ROIs were extracted for the following thalamic nuclei: Vc and Vim. The centroid of these ROIs was com pared to the location of the non-rigid (thresholded) CDRs for S1 and M1. In this case, the difference between a CDR and the tract itself is t hat the CDR only includes the parts of the tract that is within the seed region (thalamus in this case). The location of the centroid of various regions were compared against other vari ous ROIs. The following differences in centroid location were examined with fiber in tensity thresholds noted in parentheses: S1 (at 800) vs. Vc, S1 (at 1000) vs. Vc, M1 (at 800) vs. Vim, M1 (at 1000) vs. Vim, Vc vs. Vim, S1 (at 800) vs. M1 (at 800), and (Vc vs. Vim) vs. (S1 (at 800) vs M1 (at 800)). The final difference was used to examine the spatial relationship between atlas sensory and motor areas versus the analogous relations hip between tract based CDRs for sensory and motor. Differences between locations of the centro id of various regions are tabulated in Table 4-1. The centroids from both the S1 and M1 CDRs do not appear to change significantly in location when threshold ed from 800 to 1000. The mean difference in centroid location for both S1 and M1 CDRs only changed by 0.02 mm (in any given direction) as the threshold changed. Si nce both 800 and 1000 provided similar coverage of MER cells, the amount of coverage of the CDR in this threshold range does not appear to change significantly. The difference in the location of the centroid of the tract based CDR to its analogous atlas structure (S1 to Vc and M1 to Vim) ranged from about 79

PAGE 80

1.5 mm to 3 mm along any particular axis. T he final difference examined (Vc vs. Vim versus S1 vs. M1) had differences of less than 1mm for both the AP and LT axes but a difference of -2.8 mm for the AX co mponent. The atlas regions show some correspondence with tract based CDRs, but the centroids of the CDRs are not highly aligned with atlas structures. The relative lo cation of the centroids of sensory and motor CDRs appear to have a spatial relationship that is similar to that of the Vc and Vim atlas structures. Discussion and Conclusions Diffusion tractography of cortical areas to the thalamus has previously been observed to correlate with various thalamic nuclei (Behrens et al., 2003a). Since the thalamus has connections to numerous co rtical areas, tractography between the thalamus to various cortical areas of pres umably known function to may be useful to classify thalamic areas of func tion. In this chapter, voxels in the thalamus were used as seeds for tracts that went into S1 somato sensory cortex and M1 primary motor cortex. Behrens et al (Behrens et al., 2003a) has prev iously classified voxels as belonging to a particular tractography connected cortical area solely by which cortical region had the highest fiber intensity associated with it (ri gid definition). This definition may be limiting since voxels may have fibers from multiple trac ts that inhabit it especially if the voxels are of larger size. An alternative definit ion that has been previously proposed is to consider that each voxel may have contribut ions from various fibers and thereby not assigning designations solely based off maxi mum fiber intensity (non-rigid definition). MER data that was acquired from each patient allowed for direct comparison with tractography results to validate both of the definitions discu ssed previously. In the rigid definition, the location of both sensory and motor responsive MER cells was found to lie 80

PAGE 81

within the M1 CDR (with the exception of one sensory responsive cell). Rigidly defined CDRs grossly overestimated t he extent of the M1 CDR and underestimate the extent of the S1 CDR. The non-rigid definition for S1 and M1 CDRs allows for mixing for S1 and M1 tracts within a voxel. If a threshold fiber intensity is considered in the range of 800 to 1000 (relative to a starting seed count of 5000 per voxel), the S1 CDR can be used to define a region that includes a majority of sensory responsive cells while limiting the number of motor responsive cells within it. It should be noted th at no equivalent threshold was found for the M1 CDR. This di screpancy may be due to the distribution of fibers for both S1 and M1 tracts. The more anterior components of the tract tend to have a lower fiber intensity than the more posteri or parts. This effect can be seen in Figure 4-3B where the more anterior portion of the M1 tract shows a less intense fiber intensity (dull shade of red) versus the more intense fi ber intensity of the pos terior portion (bright orange color). A possible explanation for this phenomenon may be related to anterior components of the tract in t he thalamus being linked with more lateral components of the cortex (Figure 4-4). More lateral fibers are difficult to accurately model due to the necessity of modeling multiple fibers for the br anching of medial tracts to lateral cortical areas (Behrens et al., 2007). The diffusion weighted acquisition used in this study acquired only 12 distinct directions which may be able to model multiple fibers with sufficient signal to noise rati o, but more directions may help reduce this effect. This model limitation may also contribute to poor correlation of rigidly defined CDRs since the boundary of interest is bet ween the anterior portions the S1 tract and the posterior portions of the M1 tract. 81

PAGE 82

CDRs for sensory and motor areas of the thalamus show some correspondence with analogous atlas regions. In general, the m ean difference in centroid location of a CDR to its analogous atlas structure was between about 1.5 to 3 mm corresponding to roughly one to two voxel dimensions for the diffusion weighted scans. While this correspondence is not highly accurate, the relative relationship between sensory and motor CDRs compared with the analogous regions in the atlas shows a better correspondence in the AP and LT components. Since the CDRs are created from tracts, the superior to inferior direction may not co rrespond as well since the tracts in that region of the brain run generally from superior to inferior. In the atlas t here are dorsal atlas structures that are superior to bot h Vc and Vim but since the CDRs are created from tracts that may be running through those dorsal nucle i of the thalamus, the AX component of the centroid of the CDRs may not be accurate. Overall, the non-rigid classification of CDRs appears to provide a reasonable description of sensory and motor portions of the thalamus. These areas do not necessarily perfectly align with atlas pr edicted structures but do show some correspondence in their relative spatial a rrangement. MER electrophysiology does not always align with the prediced Vim/Vc atlas boundaries as well. Cellular MER data from patients fit well with the CDR regions for bot h sensory and motor if a proper threshold was chosen. In the case of the particular data that was examined, this threshold could reasonably be set to between 800 and 1000 (using a seed of 5000 per voxel) with minimal change in the centroid of the CDR wit hin that range. In DBS since the front boundary of Vc is used as a landmark for impl antation, the ability to preoperatively directly determine the location of sensory thalamus may provide additional information 82

PAGE 83

that will be useful for thalamic targeting in essential tremor patients. Since a large number of patients was not feasible for this study, futu re work examining this method across a large set of patients may help to determine the overall accuracy of diffusion tractography towards targeting in DBS. 83

PAGE 84

Table 4-1. Mean differences between centro ids of various regions in millimeters. AP LT AX S1 vs Vc at 800 1.44 1.11 1.92 1.02 2.86 1.08 S1 vs Vc at 1000 1.42 1.14 1.94 0.99 2.86 1.08 M1 vs Vim at 800 2.30 0.89 2.62 1.01 0.06 1.45 M1 vs Vim at 1000 2.28 0.92 2.64 0.98 0.06 1.45 Vc vs Vim 3.28 0.63 1.60 0.16 -4.28 0.29 S1 vs M1 at 800 2.42 0.72 0.90 0.19 -1.48 0.34 (Vc vs Vim) vs (S1 vs M1) at 800 0.86 0.78 0.70 0.14 -2.80 0.58 Errors are standard deviations. 84

PAGE 85

Figure 4-1. Diagram of ROIs used in th is study. Blue represents S1 somatosensory cortex, red represents M1 primary motor cortex, green represents SMA supplementary motor area, and yellow represents pre-SMA presupplementary motor area. 85

PAGE 86

A 0.00% 10.00% 20.00% 30.00% 40.00% 50.00% 60.00% 70.00% 80.00% 90.00% 100.00% 02004006008001000120014001600180020002200240026002800300032003400360038004000 Fiber Intensity S1 Sensory S1 Passive Motor B 0.00% 10.00% 20.00% 30.00% 40.00% 50.00% 60.00% 70.00% 80.00% 90.00% 100.00%0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 6500 7000 7500 8000 8500 9000 9500 10000 10500 11000 11500 12000 12500 13000 13500 14000 14500 15000 15500 16000 MoreFiber Intensity M1 Sensory M1 Passive Motor Figure 4-2. Cumulative histogr am of sensory and passive mo tor responsive cells that correspond to particular fiber intens ities in A) S1 and B) M1 tracts. 86

PAGE 87

A B Figure 4-3. Exemplar cases of A) S1 and B) M1 tracts from the thalamus at a threshold fiber intensity of 800 (4000 max intensity). 87

PAGE 88

A B Figure 4-4. Exemplar case of tracts seeded fr om the thalamus to different portions of the M1 primary motor cortex. Cortical ROIs for tracts are shown in panel A with respective tracts (by colors) in t he thalamus shown in panel B. The more medial portions of M1 have connections to more posterior portions of the thalamus while more lateral portions of M1 have more anterior connections. 88

PAGE 89

CHAPTER 5 A SYSTEM FOR MULTI-MODAL IMAGE GUIDANCE Aims Our aim was to prove (through proof of conc ept) that the methods developed in this work could be implemented on a clinically f easible manner towards the targeting of deep brain stimulati on (DBS) surgery. Methods and Materials Software All in-house code was created using Matl ab. Software was created to integrate multiple imaging techniques towards targeting for DBS. This integration was accomplished through the creation of software t hat can read in multiple types of imaging data. The simplest type of imaging data t hat the software can read and display are anatomic image sets such as single volu me CT, T1w MP-RAGE, T2w FLAIR, and FGATIR. The software can also be used to import a non-primary (secondary) data such as functional MRI or diffusion we ighted. In the case of t he secondary dataset, it is necessary to fuse these data into the space of the primary ( anatomic) image set. Resampling of any data datasets (primary or secondary) along different planes for either the purposes of viewing or fusion occur on-t he-fly in a single step process whereby all transformation matrices are multiplied together prior to resampling. This resampling procedure reduces interpolation errors that would occur if interpolation were used to resample the dataset at each step in the process. Fusion specifically occurs in the software using manual methods. A base secondary image can be loaded that is in the same space as the processed data (ie. probabilistic tractography fiber intensity data or 89

PAGE 90

statistical functional maps) but still has enough anatomic features to identify the underlying brain anatomy. An example of a base secondary image set for diffusion data would be the non-diffusion weighted (b =0) image. The base secondary image is then contoured at various intensity levels t hat are user adjusted and overlaid on the primary anatomic image (Figur e 5-1). Six degrees of freedom (3 translation and 3 rotation) are used to adjust the contours of the base secondary image to align the secondary image to the space of t he primary anatomic image space. Following fusion, processed (secondary) data can be loaded into the software and overlaid on the anatomic im age. The software does not distinguish between functional statistical maps, probabilistic tractography m aps or other data so t he user must provide reasonable values to the software for the co lor scaling of the secondary dataset. Color scaling can be chosen using two adjustments, threshold intensity and max intensity. Colors are scaled between the two values resu lting in an overlay on the anatomic image that varies in color from dull red to bright orange as the secondary (processed) image voxel intensity varies from threshold to max in tensity, respectively. In the special case of diffusion tensor image data, a colorized fr actional anisotropy image that shows the directionality of t he primary eigenvector as a colo r can also be loaded and viewed. For targeting purposes, additional views can be used reorient the image data (both primary and secondary) along probe views. The standard orientation view is in image space (typically coordinates assigned from the scanner) or stereot actic space if the image data has already been fused to stereotact ic space. The images can be aligned to an anatomic coordinate system gi ven by the AC/PC points. Finally, a track view can be used to view image data along para-sagittal and para-coronal views that show slices 90

PAGE 91

oriented along the sagittal and coronal plane s angled to fit the probe trajectory. MER maps and atlas data can be loaded into the software and displayed as has been discussed in previous chapters. Image Acquisition and Processing Imaging data from one exempl ar patient with essential tremor was used. This patient was screened and found to be a DBS c andidate as previously described. All imaging methods previously discussed were acquired on this patient including: stereotactic CT, T1w MP-RAGE, T2w FLAIR, T1w FGATIR, and 12-direction diffusion weighted images. All scans acquired resulted in a protocol that was less than one hour of clinical scanner time. The diffusion data we re processed as described in chapter 4. Multiple ROIs were extracted for tractogr aphy including thalamus, S1 somatosensory cortex, and M1 primary motor cortex. Tract ography from seeds within the thalamus was completed on S1 and M1 to create two tracts representing sensory and motor thalamus, respectively. Secondary data (diffusion weig hted) were fused to the FGATIR images and tracts were displayed as an overlay. Results Imaging data from the exempl ar patient was used to demonstrate integration of all previously mentioned imaging tec hniques into the clinical targeting software. An ET patients imaging data was ac quired and processed as described. These data were displayed using the developed software at the c linically chosen target. Figure 5-2A is an anatomic view of the FGATIR with atlas overlays from the software. Once a target and entry point has been chosen (or target point and entry angles), a track view can be used to visualize the imaging data along potentia l probe planes. Figure 5-2B shows a track view of FGATIR data with atlas overlays. Vi ews along the trajectory of a potentially 91

PAGE 92

implanted electrode allow for vi sualization of both the targ et and non-target structures that may be within stimulating distances. Secondary imaging data (diffusion tractography) was overlaid and shown in tr ack views in Figure 5-3A and 5-3B for sensory and motor tracks, respectively. Figure 5-3A shows a tract (thresholded at 800) from the thalamus to S1 al ong a potential target trajectory. Figure 5-3B shows a tract (thresholded at 800) from the thalamus to M1 along the same target trajectory. Discussion and Conclusions Previous chapters have shown potential fo r clinical utility of advanced imaging methods towards targeting in DBS but have not directly addressed clinic al feasibility. In this chapter, all of the pr eviously described imaging methods were acquired on an exemplar patient during the c ourse of the patients standard clinical pre-DBS MRI scan. These imaging methods were integrated into software that was designed in-house to provide a clinical platform for these advanced imaging modalities. Fu ll integration of all of these methods into software that allows for display and targeting based off any (and potentially all) of these m odalities has not previously been reported. The clinical acquisition and implementation of all of the methods provides a proof of concept for clinical feasible multi-modal targeting us ing methods described in previous chapters. Simple clinical implementation of diffusion tractography may be achievable through streamlining of the pr ocedure used to create ROIs. In the current methodology, someone with significant neuroanatomy knowledge must create ROIs for thalamus (semi-manually), S1 (manually), and M1 (m anually). The process for semi-manual ROI creation of the thalam us requires a user to deform the atlas to fit the anatomy of the patients brain. S1 and M1 delineation, in the procedure used for this work, required completely manual hand drawn outlines of these cortical areas. A potential for 92

PAGE 93

automation would be to use segmentation software for thalamus, S1, and M1. For example, the FSL software package can pot entially automatically segment out the thalamus as well as these cortical areas. While care must be used to ensure that segmentation is completed successfully, these methods may at the least provide a good first start towards automated segmentati on. Full automation of segmentation and processing techniques would streamline th is process and allow these methods to become more viable as a standard clinic al method for targeting in DBS. 93

PAGE 94

Figure 5-1. Diffusion weighted secondary (b = 0) images fused to an FGATIR primary image. Green contours represent various isointensity contours of the diffusion weighted image onto anatomic (primary) space. 94

PAGE 95

A B Figure 5-2. FGATIR primary images overlaid with atlas cont ours in A) anatomic views and B) track trajectory views. The r ed dotted line in panel B represents the planned track trajectory along the oblique views. 95

PAGE 96

A B Figure 5-3. FGATIR primary images overlaid with atlas contours and A) S1 and B) M1 tracks thresholded at 800 and 5000, res pectively. The dotted line represents a possible target trajectory. 96

PAGE 97

CHAPTER 6 OVERALL CONTRIBUTIONS AND FUTURE WORK In the previous chapters, several imaging guidance methods were developed, evaluated, and integrated into a clinical mu lti-modal image guidanc e platform that was created for this work. Two imaging methods we re specifically evaluated: the FGATIR scan and probabilistic diffusion tractography (PDT ). A novel protocol, the FGATIR scan, was developed in this work by modifying the inversion time of a clinical 3D MP-RAGE sequence to nullify the white matter signal. The FGATIR scan was found to have enhanced contrast of the STN, VL thalam us, and GPi versus respective surrounding structures when compared with standard T1w MP-RAGE and T2w FLAIR. The FGATIR scan was further evaluated for clinical utility through a targeting st udy which compared targeting based off multiple methods to det ermine if targeting using the FGATIR scan significantly helped with targeting in DBS. Four methods were specifically examined: CT only, CT+T1w MP-RAGE, CT+T1w MP-RAGE+Atlas, and CT+T1w MPRAGE+Atlas+FGATIR. This study revealed that targeting using the fourth method (which included the FGATIR) was not signific antly better than more basic methods of targeting for the specific target studied, t he STN. A potential reason that this method was not significantly more useful is that the target studied, the STN, is a target that may be rather consistently targeted based off known distances from commonly used anatomic landmarks alone. While targeting of the STN may not benefit from using the enhanced methods developed, other targets that are more variable in size and location relative to central anatomic landmarks may stand to benefit from these methods. Future work to evaluate targeting using the FG ATIR scan may help to determine if enhanced targeting using the FGATIR scan is useful in various current and new targets of DBS. 97

PAGE 98

The enhanced contrast of the FGATIR sequence may also help to establish criteria to drive an automated non-rigid deformation of a standard anatomic atlas to an individual patient anatomy. A study was also conducted to evaluat e PDT delineated regions (connectivity defined regions or CDRs) for motor and sens ory within the thalamus. These CDRs were created by using the FSL software package to create probabilistic diffusion maps to create the following tracts: t halamus to the S1 primary somatosensory cortex and thalamus to the M1 primary motor cortex. Using the PDT re sults, a threshold for fiber intensity of the PDT was determined that showed correlation with sensory and motor microelectrode data. These CDRs for sensor y and motor thalamus were also found to spatially correlate with thalamic nuclei (f rom atlas structures) within a couple of millimeters. Finally, an in-house developed software and hardware system was created that integrated all of these methods (as well as standard imaging methods) into a single multi-modal platform. This system allows fo r clinical targeting based off of all these methods combined together. All of the scans used within this work were included as part of the standard clinical pre-DBS MRI protocol and fit within the one hour timeframe necessary for scans to be practical at the clinic al level. The end result of this work is a clinically feasible methodology for multi-mo dal targeting of DBS integrating multiple novel techniques. The methods developed in this work were shown to provide enhanced structural contrast (using the FGATIR scan) and func tional contrast (using PDT to delineate CDRs) compared with standard imaging techniqu es. The methods developed in this 98

PAGE 99

work were validated using standard targets of DBS to show potential for clinical utility. Although, these methods may not offer signifi cant benefit for current targets, novel targets of DBS (where the location may vary significantly across patients) may provide contrast to localize these regions where little to none previously existed. Novel targets for DBS such as the centromedian nucleus of the thalamus for Tourettes syndrome, anterior limb of the internal capsule for obse ssive compulsive disorder, and many others may provide the opportunity in the future to clinically test the developed framework in providing the necessary contrast to successf ully target and potentially help reduce the symptoms of these and many ot her neurological disorders. 99

PAGE 100

LIST OF REFERENCES Albin, R.L., Young, A.B., Penney, J.B., 1989. The functional anatomy of basal ganglia disorders. Trends Ne urosci. 12, 366. Alexander, G.E., Crutcher, M. D., DeLong, M.R., 1990. Basa l ganglia-thalamocortical circuits: parallel substrates for moto r, oculomotor, "prefrontal" and "limbic" functions. Prog. Brain Res. 85, 119. Alexander, G.E., DeLong, M.R., Strick, P.L., 1986. Parallel organization of functionally segregated circuits linking basal ganglia and cortex. Annu. Rev. Neurosci. 9, 357. Amirnovin, R., Williams, Z.M., Cosgrove, G. R., Eskandar, E.N., 2006. Experience with microelectrode guided subthalamic nucleus deep brain stimulation. Neurosurgery 58, ONS96; discussion ONS196-102. Aravamuthan, B.R., Stein, J.F., Aziz, T.Z. 2008. The anatomy and localization of the pedunculopontine nucleus determined using probabilistic diffusion tractography [corrected]. Br. J. Neurosurg. 22 Suppl 1, S25. Barbas, H., Henion, T.H., Dermon, C.R., 1991. Diverse thalamic projections to the prefrontal cortex in the rhesus monkey. J. Comp. Neurol. 313, 65. Basser, P.J., Mattiello, J., LeBihan, D., 1994. MR diffusion tensor spectroscopy and imaging. Biophys. J 66, 259. Behrens, T.E., Berg, H.J., Jbabdi, S., Ru shworth, M.F., Wool rich, M.W., 2007. Probabilistic diffusion tractogr aphy with multiple fibre or ientations: What can we gain? Neuroimage 34, 144. Behrens, T.E., Johansen-Berg, H., Woolrich M.W., Smith, S.M ., Wheeler-Kingshott, C.A., Boulby, P.A., Barker, G.J., Siller y, E.L., Sheehan, K., Ciccarelli, O., Thompson, A.J., Brady, J.M., Matthews, P.M., 2003a. Non-invasive mapping of connections between human thalamus and co rtex using diffusion imaging. Nat. Neurosci. 6, 750. Behrens, T.E., Woolrich, M.W. Jenkinson, M., Johansen-Berg, H., Nunes, R.G., Clare, S., Matthews, P.M., Brady J.M., Smith, S.M., 2003b. Characterization and propagation of uncertainty in diffusion-we ighted MR imaging. Magn. Reson. Med. 50, 1077. Benabid, A.L., Pollak, P., Gervason, C., Ho ffmann, D., Gao, D.M., Hommel, M., Perret, J.E., de Rougemont, J., 1991. Long-term suppression of tremor by chronic stimulation of the ventral intermediat e thalamic nucleus. Lancet 337, 403. 100

PAGE 101

Benabid, A.L., Pollak, P., Ho mmel, M., Gaio, J. M., de Rougemont, J., Perret, J., 1989. Treatment of Parkinson trem or by chronic stimulation of the ventral intermediate nucleus of the thalamus. Rev. Neurol. (Paris) 145, 320. Beurrier, C., Bioulac, B., Audin, J., Ha mmond, C., 2001. High-fr equency stimulation produces a transient blockade of voltagegated currents in subthalamic neurons. J. Neurophysiol. 85, 1351. Bootin, M.L., 2006. Deep brain stimulati on: overview and update. J. Clin. Monit. Comput. 20, 341. Butson, C.R., McIntyre, C.C., 2006. Role of electrode design on the volume of tissue activated during deep brai n stimulation. J. N eural. Eng. 3, 1. Carpenter, M.B., Carleton, S. C., Keller, J.T., Conte, P ., 1981. Connections of the subthalamic nucleus in the monkey. Brain Res. 224, 1. Chakravarty, M.M., Sadikot, A.F., Germann, J., Bertrand, G., Collins, D.L., 2008. Towards a validation of atlas warp ing techniques. Med. Image. Anal. Chen, S.Y., Lee, C.C., Lin, S. H., Hsin, Y.L., Lee, T.W., Yen, P.S., Chou, Y.C., Lee, C.W., Annie Hsieh, W., Su, C.F., Lin, S.Z., 2006. Mi croelectrode recording can be a good adjunct in magnetic resonance image-directed subthalamic nucleus deep brain stimulation for parkinsonism. Surg. Neurol. 65, 253-260; discussion 260. Chenevert, T.L., Brunberg, J.A., Pipe, J.G., 1990. Anisot ropic diffusion in human white matter: demonstration with MR techni ques in vivo. Radiology 177, 401. Cosyns, P., Gabriels, L., Nuttin, B., 2003. Deep brain stimulation in treatment refractory obsessive compulsive disorder. Verh K. Acad. Geneeskd. Belg. 65, 385; discussion 399. Cuny, E., Guehl, D., Burbaud, P., Gross, C., Dousset, V ., Rougier, A., 2002. Lack of agreement between direct magnetic resonance imaging and statistical determination of a subthalamic target: the role of electrophysiological guidance. J. Neurosurg. 97, 591. DBSPDSG, 2001. Deep-brain stimul ation of the subthalamic nucleus or the pars interna of the globus pallidus in Parkinson's di sease. N. Engl. J. Med. 345, 956. Deoni, S.C., Josseau, M.J., Ru tt, B.K., Peters, T.M., 2005a. Visualization of thalamic nuclei on high resolution, multi-averaged T1 and T2 ma ps acquired at 1.5 T. Hum. Brain Mapp. 25(3), 353. 101

PAGE 102

Deoni, S.C., Peters, T.M., Rutt, B.K., 2005b. High-resolution T1 and T2 mapping of the brain in a clinically acceptable ti me with DESPOT1 and DESPOT2. Magn. Reson. Med. 53(1), 237. DeVito, J.L., Anderson, M.E., 1982. An autor adiographic study of efferent connections of the globus pallidus in Macaca mulatta. Exp. Brain Res. 46, 107. Doran, M., Hajnal, J.V., V an Bruggen, N., King, M.D., Young, I.R., Bydder, G.M., 1990. Normal and abnormal white matter tracts shown by MR imaging using directional diffusion weighted sequences. J. Com put. Assist. Tomogr. 14, 865. Dormont, D., Ricciardi, K.G., Tande, D., Para in, K., Menuel, C., Ga lanaud, D., Navarro, S., Cornu, P., Agid, Y., Yelnik, J., 2004. Is the subthalamic nucleus hypointense on T2-weighted images? A correlation st udy using MR imaging and stereotactic atlas data. AJNR Am. J. Neuroradiol. 25, 1516. Dostrovsky, J.O., Levy, R., Wu, J.P., Hutchi son, W.D., Tasker, R.R., Lozano, A.M., 2000. Microstimulation-induced inhibition of neuronal firing in human globus pallidus. J. Neurophysiol. 84, 570. Duffner, F., Schiffbauer, H., Breit, S., Fr iese, S., Freudenstein, D., 2002. Relevance of image fusion for target point determinat ion in functional neurosurgery. Acta Neurochir. (Wien) 144, 445. Durif, F., Lemaire, J.J., Debilly, B., Dor dain, G., 1999. Acute and chronic effects of anteromedial globus pallidus stimulation in Parkinson's disease. J. Neurol. Neurosurg. Psychiatry 67, 315. Filali, M., Hutchison, W.D., Palter, V.N., Lozano, A.M., Do strovsky, J.O., 2004. Stimulation-induced inhibiti on of neuronal firing in human subthalamic nucleus. Exp. Brain Res. 156, 274. Francois, C., Percheron, G., Yelnik, J., Tande, D., 1988. A topographic study of the course of nigral axons and of the distri bution of pallidal axonal endings in the centre median-parafascicular comple x of macaques. Brain Res. 473, 181. Francois, C., Yelnik, J., Percheron, G., Tand e, D., 1994. Calbindin D-28k as a marker for the associative cortical territory of the striatum in macaque. Brain Res. 633, 331. Franzini, A., Ferroli, P., Leone, M., Broggi, G., 2003. Stimulation of the posterior hypothalamus for treatment of chroni c intractable cluster headaches: first reported series. Neurosurgery 52, 1095; discussion 1099. 102

PAGE 103

Greenberg, B.D., Malone, D.A., Friehs, G.M., Rezai, A.R., Kubu, C.S., Malloy, P.F., Salloway, S.P., Okun, M.S., Goodman, W.K., Rasmussen, S.A., 2006. Threeyear outcomes in deep brain stimulat ion for highly resistant obsessivecompulsive disorder. Neur opsychopharmacology 31, 2384. Grill, W.M., Snyder, A.N., Miocinovic, S ., 2004. Deep brain stimulation creates an informational lesion of the stimulat ed nucleus. Neuroreport 15, 1137. Gross, R.E., Lombardi, W.J., Lang, A.E., Duff, J., Hutchison, W.D., Saint-Cyr, J.A., Tasker, R.R., Lozano, A.M., 1999. Relations hip of lesion location to clinical outcome following microelectrode-guided pa llidotomy for Parkinson's disease. Brain 122 ( Pt 3), 405. Gutman, D.A., Holtzheimer, P.E., Behrens T.E., Johansen-Berg, H., Mayberg, H.S., 2009. A tractography analysis of two deep brain stimulat ion white matter targets for depression. Biol. Psychiatry 65, 276. Hamel, W., Fietzek, U., Morsnowski, A., Sc hrader, B., Herzog, J., Weinert, D., Pfister, G., Muller, D., Volkmann, J., Deusch l, G., Mehdorn, H.M., 2003. Deep brain stimulation of the subthalamic nucleus in Parkinson's disease: evaluation of active electrode contacts. J. Neurol Neurosurg. Psychiatry 74, 1036. Hashimoto, T., Elder, C.M., Ok un, M.S., Patrick, S.K., Vite k, J.L., 2003. Stimulation of the subthalamic nucleus changes the fi ring pattern of pallidal neurons. J. Neurosci. 23, 1916. Hirai, T., Jones, E.G., 1989. A new parcellation of the human thalamus on the basis of histochemical staining. Brain Re s. Brain Res. Rev. 14, 1. Hung, S.W., Hamani, C., Lozano, A.M., Poon, Y.Y., Piboolnurak, P., Miyasaki, J.M., Lang, A.E., Dostrovsky, J.O., Hutchis on, W.D., Moro, E., 2007. Long-term outcome of bilateral pallidal deep brain stimulation for primary cervical dystonia. Neurology 68, 457. Hutchison, W.D., Allan, R.J. Opitz, H., Levy, R., Dost rovsky, J.O., Lang, A.E., Lozano, A.M., 1998. Neurophysiological identificat ion of the subthalamic nucleus in surgery for Parkinson's dis ease. Ann. Neurol. 44, 622. Karachi, C., Francois, C., Pa rain, K., Bardinet, E., Tande, D., Hirsch, E., Yelnik, J., 2002. Three-dimensional cartography of functional territories in the human striatopallidal complex by using calbin din immunoreactivity. J. Comp. Neurol. 450, 122. Karachi, C., Yelnik, J., Tande, D., Trembl ay, L., Hirsch, E.C., Fr ancois, C., 2005. The pallidosubthalamic projection: an anatomical substrate for nonmot or functions of the subthalamic nucleus in prim ates. Mov. Disord. 20, 172. 103

PAGE 104

Kelly, R.M., Strick, P.L., 2004. Macro-architecture of ba sal ganglia loops with the cerebral cortex: use of rabies virus to reveal multisynaptic circuits. Prog. Brain Res. 143, 449. Kitajima, M., Korogi, Y., Kak eda, S., Moriya, J., Ohnari, N ., Sato, T., Hayashida, Y., Hirai, T., Okuda, T., Yamashita, Y ., 2008. Human subthalamic nucleus: evaluation with high-resolution MR imagi ng at 3.0 T. Neuror adiology 50(8), 675 681. Koller, W., Pahwa, R., Busenbark, K., Hubble, J., Wilkinson, S., Lang, A., Tuite, P., Sime, E., Lazano, A., Hauser, R., Malapira T., Smith, D., Ta rsy, D., Miyawaki, E., Norregaard, T., Kormos, T., Olanow C.W., 1997. High-frequency unilateral thalamic stimulation in the treatment of essential and parkinsonian tremor. Ann. Neurol. 42, 292. Kovacs, N., Auer, T., Balas, I., Karadi, K., Zambo, K., Schwarcz, A., Klivenyi, P., Jokeit, H., Horvath, K., Nagy, F., Janszky J., 2009. Neuroimaging and cognitive changes during deja vu. Epilepsy Behav. 14, 190. Kumar, R., 2002. Methods for programmi ng and patient management with deep brain stimulation of the globus pallidus for the treatment of advanced Parkinson's disease and dystonia. Mov. Disord. 17 Suppl 3, S198. Kumar, R., Lozano, A.M., Si me, E., Lang, A.E., 2003. Longterm follow-up of thalamic deep brain stimulation for essential and parkinsonian tremor. Neurology 61, 1601. Lanotte, M.M., Rizzone, M., Berg amasco, B., Faccani, G., Me lcarne, A., Lopiano, L., 2002. Deep brain stimulation of t he subthalamic nucleus: anatomical, neurophysiological, and outcome correlations with the effe cts of stimulation. J. Neurol. Neurosurg. Psychiatry 72, 53. Lee, C., Young, B., Sanders, M.F., 2006. The role of the supramammillary commissure in MR localization of the subthalamic nuc leus. Stereotact. Funct. Neurosurg. 84, 193. Lee, M.W., De Salles, A.A., Fr ighetto, L., Torres, R., Behnke, E., Bronstein, J.M., 2005. Deep brain stimulation in intraoperati ve MRI environment comparison of imaging techniques and electrode fixation methods. Minim. Invasive Neurosurg. 48, 1. Lehericy, S., Bardinet, E., Tremblay, L., V an de Moortele, P.F., Po chon, J.B., Dormont, D., Kim, D.S., Yelnik, J., U gurbil, K., 2006. Motor control in basal ganglia circuits using fMRI and brain atlas approa ches. Cereb. Cortex 16, 149. 104

PAGE 105

Limousin, P., Krack, P., Po llak, P., Benazzouz, A., Ardoui n, C., Hoffmann, D., Benabid, A.L., 1998. Electrical stimulation of the subthalamic nucleus in advanced Parkinson's disease. N. Engl. J. Med. 339, 1105. Lozano, A.M., Mayberg, H.S., Giacobbe, P ., Hamani, C., Craddock, R.C., Kennedy, S.H., 2008. Subcallosal cingulate gyru s deep brain stimulation for treatmentresistant depression. Biol. Psychiatry 64, 461. McClelland, S., 3rd, Ford, B., Senatus, P.B., Winfield, L.M., Du, Y.E., Pullman, S.L., Yu, Q., Frucht, S.J., McKhann, G.M., 2nd, Goodman, R.R., 2005. Subthalamic stimulation for Parkinson disease: dete rmination of electrode location necessary for clinical efficacy. Neurosurg. Focus 19, E12. McIntyre, C.C., Grill, W.M., S herman, D.L., Thakor, N.V., 2004a. Cellular effects of deep brain stimulation: model -based analysis of activation and inhibition. J. Neurophysiol. 91, 1457. McIntyre, C.C., Mori S., Sherman, D.L., Thakor, N.V., Vitek, J.L., 2004b. Electric field and stimulating influence generated by deep brain stimulation of the subthalamic nucleus. Clin. Neurophysiol. 115, 589. McIntyre, C.C., Savasta, M., Kerkerian-Le Go ff, L., Vitek, J.L., 2004c. Uncovering the mechanism(s) of action of deep brain stimulat ion: activation, inhibition, or both. Clin. Neurophysiol. 115, 1239. Mercado, R., Mandat, T., Moore, G.R., Li, D., MacKay, A ., Honey, C.R., 2006. Threetesla magnetic resonance imaging of the ventrolateral thalamus: a correlative anatomical description. J. Neurosurg. 105(2), 279. Middleton, F.A., Strick, P.L., 2000. Basal ganglia and cerebellar loops: motor and cognitive circuits. Brain Res. Brain Res. Rev 31, 236. Mink, J.W., 1996. The basal ganglia: focused selection and inhibition of competing motor programs. Prog Neurobiol. 50, 381. Mink, J.W., 2003. The Basal Ganglia and involuntary movements: impa ired inhibition of competing motor patterns. Arch. Neurol. 60, 1365. Moseley, M.E., Cohen, Y., Kucharczyk, J., Mintorovitch, J., As gari, H.S., Wendland, M.F., Tsuruda, J., Norman, D., 1990a. Diffusion-weighted MR imaging of anisotropic water diffusion in cat cent ral nervous system. Radiology 176, 439 445. 105

PAGE 106

Moseley, M.E., Kucharczyk, J., Mintorovitch J., Cohen, Y., Kurhanewicz, J., Derugin, N., Asgari, H., Norman, D., 1990b. Diffu sion-weighted MR im aging of acute stroke: correlation with T2-weighted and magnetic susceptib ility-enhanced MR imaging in cats. AJNR Am. J. Neuroradiol. 11, 423. Nakano, K., Hasegawa, Y., Tokushige, A., Nakagawa, S., Kayahara, T., Mizuno, N., 1990. Topographical projecti ons from the thalamus, subthalamic nucleus and pedunculopontine tegmental nucleus to the striatum in the Japanese monkey, Macaca fuscata. Brain Res. 537, 54. Nauta, H.J., Cole, M., 1978. Efferent projections of the subthalamic nucleus: an autoradiographic study in monkey and cat. J. Comp. Neurol. 180, 1. Nowinski, W.L., Belov, D., Pollak, P., Benabi d, A.L., 2005. Statistical analysis of 168 bilateral subthalamic nucleus implant ations by means of the probabilistic functional atlas. Neurosurgery 57, 319-330; discussion 319. Nuttin, B., Gybels, J., Cosyn s, P., Gabriels, L., Meyerson, B., Andreewitch, S., Rasmussen, S.A., Greenberg, B., Friehs, G., Rezai, A.R., Montgomery, E., Malone, D., Fins, J.J., 2003. Deep brain stimulation for psychiatric disorders. Neurosurg. Clin. N. Am. 14, xvxvi. Nuttin BJ, Gabriels L, van Kuyck K, Cosyns P, 2003. Electrical stimulation of the anterior limbs of the internal capsul es in patients with severe obsessivecompulsive disorder: anecdotal reports Neurosurg. Clin. N. Am. 14, 267. Okun, M.S., Bowers, D., Springer U., Shapira, N.A., Malone, D ., Rezai, A.R., Nuttin, B., Heilman, K.M., Morecraft, R.J., Rasmussen, S.A., Gr eenberg, B.D., Foote, K.D., Goodman, W.K., 2004a. What's in a "smile?" Intra-operative observations of contralateral smiles induced by deep brain stimulation. N eurocase 10, 271. Okun, M.S., Fernandez, H.H., Pedraza, O., Misra, M., Lyons, K.E., Pahwa, R., Tarsy, D., Scollins, L., Corapi, K., Friehs, G.M. Grace, J., Romrell, J., Foote, K.D., 2004b. Development and init ial validation of a screeni ng tool for Parkinson disease surgical candidates. Neurology 63, 161. Okun, M.S., Mann, G., Foote, K.D., Shapira, N.A., Bowers, D., Springer, U., Knight, W., Martin, P., Goodman, W.K., 2007. Deep brain stimulation in the internal capsule and nucleus accumbens region: responses observed during active and sham programming. J. Neurol. Neurosurg. Psychiatry 78, 310. Okun, M.S., Tagliati, M., Pourfar, M., Fer nandez, H.H., Rodriguez, R. L., Alterman, R.L., Foote, K.D., 2005. Management of referr ed deep brain stimulation failures: a retrospective analysis from 2 movement disorders centers. Arch. Neurol. 62, 1250. 106

PAGE 107

Okun, M.S., Vitek, J.L., 2004. Lesion ther apy for Parkinson's disease and other movement disorders: updat e and controversies. Mov. Disord. 19, 375. Ondo, W., Jankovic, J., Schwartz, K., Alm aguer, M., Simpson, R.K., 1998. Unilateral thalamic deep brain stimulation for refr actory essential tremor and Parkinson's disease tremor. Neurology 51, 1063. Owen, S.L., Heath, J., Kringel bach, M.L., Stein, J.F., Aziz T.Z., 2007. Preoperative DTI and probabilistic tractography in an amput ee with deep brain stimulation for lower limb stump pain. Br. J. Neurosurg. 21, 485. Pahwa, R., Lyons, K.E., Wilkinson, S.B., Si mpson, R.K., Jr., Ondo, W.G., Tarsy, D., Norregaard, T., Hubble, J.P., Smith, D. A., Hauser, R.A., Jankovic, J., 2006. Long-term evaluation of deep brain stimulation of the th alamus. J. Neurosurg. 104, 506. Pahwa, R., Lyons, K.E., Wilkinson, S.B., Troste r, A.I., Overman, J., Kieltyka, J., Koller, W.C., 2001. Comparison of thalamotomy to deep brai n stimulation of the thalamus in essential trem or. Mov. Disord. 16, 140. Papavassiliou, E., Rau, G., Heath, S., A bosch, A., Barbaro, N.M., Larson, P.S., Lamborn, K., Starr, P.A., 2004. Thalamic deep brain st imulation for essential tremor: relation of lead location to outcome. Neurosurgery 54, 1120; discussion 1129. Parent, A., De Bellefeuille, L., 1982. Organization of effe rent projections from the internal segment of globus pallidus in primate as revealed by fluorescence retrograde labeling method. Brain Res. 245, 201. Parent, A., Hazrati, L.N., 1995. Functional anatomy of the basal ganglia. I. The corticobasal ganglia-thalamo-cortical loop. Br ain Res. Brain Res. Rev. 20, 91. Pinsker, M.O., Volkmann, J., Falk, D., Herzog, J., Alfke, K., Stei gerwald, F., Deuschl, G., Mehdorn, M., 2008. Electrode implant ation for deep brain stimulation in dystonia: a fast spin-echo inversion -recovery sequence technique for direct stereotactic targeting of the GPI. Z entralbl. Neurochir. 69(2), 71. Pollo, C., Vingerhoets, F., Pral ong, E., Ghika, J., Maeder, P., Meuli, R., Thiran, J.P., Villemure, J.G., 2007. Localization of elec trodes in the subthalamic nucleus on magnetic resonance imaging. J. Neurosurg. 106, 36. Rampini, P.M., Locatelli, M., Alimehmeti, R., Tamma, F., Caputo, E., Priori, A., Pesenti, A., Rohr, M., Egidi, M., 2003. Multiple sequential image-fusion and direct MRI localisation of the subthal amic nucleus for deep brain stimulation. J. Neurosurg. Sci. 47, 33. 107

PAGE 108

Reich, C.A., Hudgins, P.A., Sheppard, S.K. Starr, P.A., Bakay, R.A., 2000. A highresolution fast spin-echo inversion-recovery sequence for preoperative localization of the internal globus pallidus. AJNR Am. J. Neuroradiol. 21, 928 931. Saint-Cyr, J.A., Hoque, T., Pereira, L.C., Dostrovsky, J.O., Hutchi son, W.D., Mikulis, D.J., Abosch, A., Sime, E., Lang, A.E ., Lozano, A.M., 2002. Localization of clinically effective stimulating electr odes in the human subthalamic nucleus on magnetic resonance imaging. J. Neurosurg. 97, 1152. Schaltenbrand, G., Bailey, P., 1959. Einfhr ung in die stereotaktischen Operationen, mit einem Atlas des menschlichen Gehirns. Introduction to stereotaxis, with an atlas of the human brain. Thieme, Stuttgart. Schrader, B., Hamel, W., Weinert, D., Mehdorn, H.M., 2002. Documentation of electrode localization. Mov. Disord. 17 Suppl 3, S167. Schuurman, P.R., Bosch, D.A., Bossuyt, P.M., B onsel, G.J., van Someren, E.J., de Bie, R.M., Merkus, M.P., Speelman, J.D., 2000. A comparison of c ontinuous thalamic stimulation and thalamotomy for suppression of severe tremor. N. Engl. J. Med. 342, 461. Sedrak, M., Gorgulho, A., De Salles, A.F., Frew, A., Behnke, E., Ishida, W., Klochkov, T., Malkasian, D., 2008. The role of modern imaging modalities on deep brain stimulation targeting for mental illne ss. Acta Neurochir. Suppl. 101, 3. Slavin, K.V., Thulborn, K.R., Wess, C., Nersesyan, H., 2006. Direct visualization of the human subthalamic nucleus with 3T MR im aging. AJNR Am. J. Neuroradiol. 27, 80. Starr, P.A., Christine, C. W., Theodosopoulos, P.V., Lindsey N., Byrd, D., Mosley, A., Marks, W.J., Jr., 2002. Implantation of deep brain stimulators into the subthalamic nucleus: technical appr oach and magnetic resonance imagingverified lead locations. J. Neurosurg. 97, 370. Starr, P.A., Turner, R.S., Rau, G., Lindsey, N ., Heath, S., Volz, M., Ostrem, J.L., Marks, W.J., Jr., 2004. Microelectrode-guided implan tation of deep brain stimulators into the globus pallidus internus for dystoni a: techniques, electrode locations, and outcomes. Neurosurg. Focus 17, E4. Starr, P.A., Vitek, J.L., DeLong, M., Bakay, R.A., 1999. Magnetic resonance imagingbased stereotactic localization of the globus pallidus and subthalamic nucleus. Neurosurgery 44, 303-313; discussion 313304. 108

PAGE 109

Sudhyadhom, A., 2005. Development of an atlas based image guidance system for deep brain stimulation. Nucl ear and Radiological Engi neering. University of Florida, Gainesville. Temel, Y., Visser-Vandewalle, V., 2004. Surgery in Tourette syndrome. Mov. Disord. 19, 3. Toda, H., Sawamoto, N., Hanakawa, T., Sa iki, H., Matsumoto, S., Okumura, R., Ishikawa, M., Fukuyama, H., Hashimoto, N., 2009. A novel composite targeting method using high-field magnetic resonance imaging for subthalamic nucleus deep brain stimulation. J. Neurosurg. 111, 737. Tuch, D.S., 2004. Q-ball imaging. Magn. Reson. Med. 52, 1358. Tuch, D.S., Reese, T.G., Wiegell, M.R., Makris, N., Belliveau, J.W., Wedeen, V.J., 2002. High angular resolution diffusion imagi ng reveals intravoxel white matter fiber heterogeneity. Magn. Re son. Med. 48, 577. Visser-Vandewalle, V., Temel, Y., Boon, P ., Vreeling, F., Colle H., Hoogland, G., Groenewegen, H.J., van der Linden, C., 2003. Chronic bilateral thalamic stimulation: a new therapeutic approach in intractable T ourette syndrome. Report of three cases. J. Neurosurg. 99, 1094. Volkmann, J., Herzog, J., Kopper, F., Deuschl, G., 2002. Introduction to the programming of deep brain st imulators. Mov. Disord. 17 Suppl 3, S181. Wider, C., Pollo, C., Bloch, J., Burk hard, P.R., Vingerhoets, F.J., 2008. Long-term outcome of 50 consecutive Parkinson's disease patients treat ed with subthalamic deep brain stimulation. Parkins onism Relat. Disor d. 14, 114. Wiegell, M.R., Tuch, D.S., Larsson, H.B., Wedeen, V.J., 2003. Auto matic segmentation of thalamic nuclei from diffusion tensor magnetic resonance imaging. Neuroimage 19, 391. Yamada, K., Akazawa, K., Yuen, S., Go to, M., Matsushima, S., Takahata, A., Nakagawa, M., Mineura, K., Nishimura T., 2009. MR Imaging of Ventral Thalamic Nuclei. AJNR Am. J. Neuroradiol. Yelnik, J., 2002. Functional anat omy of the basal ganglia. Mo v. Disord. 17 Suppl 3, S15. Yelnik, J., Damier, P., Demeret, S., Gervais, D., Bardinet, E. Bejjani, B.P., Francois, C., Houeto, J.L., Arnule, I., Do rmont, D., Galanaud, D., Pidou x, B., Cornu, P., Agid, Y., 2003. Localization of stimulating electr odes in patients with Parkinson disease by using a three-dimensional atlas-magnetic resonance imaging coregistration method. J. Neurosurg. 99, 89. 109

PAGE 110

Zorzi, G., Marras, C., Nar docci, N., Franzini, A., Chia pparini, L., Maccagnano, E., Angelini, L., Caldiroli, D ., Broggi, G., 2005. Stimulati on of the globus pallidus internus for childhood-onset dyst onia. Mov. Disord. 20, 1194. 110

PAGE 111

BIOGRAPHICAL SKETCH Atchar Sudhyadhom was born on November 15, 1981 in New York City (Queens). While he has lived in many place outside of Florida throughout his life, he graduated from Gaither High School in 1999 in Tampa, Florida. He received his Bachelor of Science in physics and chemistry from the Univ ersity of Florida in May 2003 where he stayed to pursue graduate studies in medica l physics. In 2005, he earned his Master of Science in medical physics where his thes is focused on software methods for image guidance in deep brain stimulation. In 2006, he started his work towards his doctorate in medical physics. Upon finishing his graduate work, he plans to continue an academic career in medical physics primarily focusing on therapy and image guidance. 111