University of Florida | Journal of Undergraduate Research | Volume 15, Issue 3 2 | Summer 2014 1 Structural Studies of the Human Brain Stem with Diffusion Magnetic Resonance Imaging Shelby Goicochea1, Luis Colon-Perez2, David B. FitzGerlad3, 5, Thomas Mareci4 1Departments of Chemistry, 2Physics, 3Neurology, 4Biochemistry and Molecular Biology, McKnight Brain Institute, University of Florida, 5North Florida/South Georgia Veterans Health System College of Liberal Arts and Sciences, University of Florida Traumatic brain injuries (TBI) affect 1.7 million Americans a year. Approximately 75% of cases are classified as mild (mTBI). However, individuals still experience chronic symptoms; several studies have documented sleep disturbances experienced by pat ients suffering from mild TBI. The circuitry for sleep regulation is located in the brain stem and was the focus of our study. We hypothesized that TBI damages brain stem tracts, impairing neuronal communication among them. The brain stem tracts were stud ied with magneti c resonance using high angular resolution diffusion images (HARDI). Structural parameters, average diffusivity (AD) and fractional anisotropy (FA), can be calculated from HARDI data to measure the tissue organization and rate of diffusion in the brain, res pectively. Degradation of the brain stem nerve fibrous structures due to mTBI will decrease the water diffusion, resulting in lower AD and FA values in the brain stem. In this manner, these diffusion values can indicate whether a nerve bundle is intac t or compromised. Brain stem images were collected for healthy subjects and regions of interest (ROIs) were drawn in the midbrain, pons, and medulla areas, based on histological slices. Average diffusivity and FA values were calculated within these areas for th e ROIs based on the following anatomical structures: cerebral crus and aqueduct; corticospinal tract; and the pyramid of medulla. Re lative standard deviation of the AD and FA values ranged from 2 12 percent. This work will lead to a data base of normal brain stem diffusion values, which will eventually be used to compare normal subjects to mTBI patients. INTRODUCTION Traumatic brain injuries (TBI) affect 1.7 million Americans annually.1 Despite the classification of the majority of TBI cases 75 percent as mild2, mild traumatic brain injury (mTBI) contributes to chronic sleep disturbances in individuals. Studies based on subjective sleep complaints3 and objective sleep quality studies have found that mTBI patients have circadian rhythm disorders4, decreased deep sleep with awakenings lasting more than 5 minutes5, difficulty falling asleep6, and lower melatonin levels7. Although the sleep disturbances are present after mTBI, the etiology of these disturbances is not fully understood and must be researched. Sleep Circuitry The mechanisms for sleep promotion and regulation are located in the brain stem.8 The midbrain, pons, and medulla are involved in the initiation and regulation of different sleep stages.9 We hypothesize that blast or blunt trauma impairs neurological communication within the sleep regulatory network by damaging small white matter tracts within these structures, causing sleep disturbances in mTBI patients. The brain stem structure can be studied using magnetic resonance imag ing (MRI) to determine the correlation between brain stem tract damage to sleep disturbances to better understand the etiology of theses disturbances. Diffusion MRI The random translation motion of water through the brain is driven by diffusion. Water mol ecules microscopically interact with tissue structure, such as fibers and cell membranes. DWI measures the displacement distribution of water through the brain, allowing for high resolution noninvasive in vivo studies of the structural organization of tiss ues.10 The restriction of diffusion in the brain due to highly organized structures is not equivalent in all directions; therefore, DWI uses a diffusion tensor to characterize the threedimensional diffusion of water molecules.11 Scalar diffusion parameters, average diffusivity (AD) and fractional anisotropy (FA), can be calculated from the diffusion tensor to provide further structural information.12 Objective Diffusion weighted imaging of the brain stem structure allows the noni nvasive study of the etiology of sleep disturbances in patients that have suffered mild traumatic
SHEL BY GOICOCHEA, LUIS COLONPEREZ, DAVID B. FITZGERLAD, THOMAS MARECI University of Florida | Journal of Undergraduate Research | Volume 15, Issue 3 | S ummer 2014 2 brain injury. We collected high angular resolution diffusion images (HARDI) of human brain stem and drew regions of interest in the midbrain, pons, and medull a. From these regions, FA and AD values will be calculated to create a database of healthy brain stem diffusion values. SUBJECTS AND METHODS Subjects We recruited 5 healthy agematched controls who have not experienced mTBI and lacked sleep disturbances. To ensure that their sleep was not disrupted, subjective data was collected through structured interviews and objective data was collected through administering the Pittsburgh Sleep Quality Index and the Epworth Sleepiness Scale. Upon completion of the sc reening process, the healthy controls underwent DWI scans. DWI Acquisition HARDI of the brain stems of 5 healthy controls was obtained in a Philips 3T magnet with a spin echo preparation and echo planar imaging readout at 1.7 mm isotropic resolution with 6 direction diffusion weightings of 100 s/mm2 and 64 direction diffusion weightings of 1000 s/mm2. Pre processing The 64 and 6 angular direction image data sets were then combined using inhouse software written in IDL, MRI Analysis Software (MAS). The comb ined data was motioncorrected using FSLs Eddy Correct.13 MAS was used to interpolate the data to a 0.85 mm isotropic resolution. Finally, quantitative AD and FA maps were created with MAS from the HARDI acquisition data. Regions of Interest The brain st em was divided into the following three sections: the midbrain, pons, and medulla. Using the transverse plane, the upper limit for the midbrain was identified by the disappearance of the third ventricle (Fig.1, line 1) and the lower limit defined by the top of the peduncles (Fig.1. line 2). The upper limit of the pons begins at the lower limit of the midbrain, and the lower limit was defined by the beginning of the corticospinal tracts (Fig.1, line 3). The top of the medulla was defined by the beginning of the corticospinal tracts and the bottom by the bottom of the cerebellum (Fig. 1, line 4), determined in the sagittal plane. Within each region of the brain stem, the transverse slices of these landmarks (lines 1 4) were recorded. The median slice is determined and selected for the ROI drawings This procedure helped compensate for the differences in structure length between healthy subjects, leading to increased reproducibility (A. Ford, PhD, oral communication, Jan 2012). Using AD and FA maps of in vivo brain stem acquisitions, the ROIs were segmented in MAS from transverse slices following histological guidelines (Fig. 2D F). 15 For the midbrain section (Fig.1, line A), the cerebral crus (peduncle) was visualized in an FA map, while the cerebral aqueduct (periaqueduct), excluding the cerebrosp inal fluid, was drawn in an AD map (Fig. 2A). Using the pons (Fig. 1, line B) and medulla (Fig. 1, line C) sections of the brain stem, FA maps were used to segment the corticospinal tract (CST) (Fig. 2B) and pyramid of medulla (CST) (Fig. 2C), respectively RESULTS DWI brain stem in vivo scans were obtained from the subjects. Due to the anisotropy in the brain originating from the degree of myelinated axonal fibers, a diffusion tensor () is necessary to properly characterize the diffusion of water in thre e dimensions given by equation 1, = where D represents the diffusion coefficient in the direction indicated by the subscript.16 The diagonalization of the diffusion tensor produces orientations and associated diffusivitivies.17 Scalar diffusion parameters, AD and FA, can be calculated (Table 1) from these eigenvalues18 after accurate segmentation of Figure 1 Human brain stem separated into midbrain, pons, and me dulla, labeled as A, B, C, respectively (including the ROI placement) edited image. 15 (Eq. 1)
University of Florida | Journal of Undergraduate Research | Volume 15, Issue 3 2 | Summer 2014 3 Table 1. AD and FA values for the left (L) and right (R) ROIs in the midbrain, pons, and medulla of the brain stem. Area ROI 1 2 3 4 5 Average RSD Average Diffusivity (AD) MidBrain peduncleL 6.93E -04 7.84E -04 6.59E -04 8.48E -04 7.29E -04 7.43E -04 9.00% peduncleR 6.92E -04 8.18E -04 7.66E -04 8.55E -04 7.48E -04 7.76E -04 7.20% periaqueductL 7.79E -04 8.38E -04 8.19E -04 8.24E -04 8.57E -04 8.24E -04 3.10% periaqueductR 7.63E -04 8.12E -04 7.93E -04 7.88E -04 8.26E -04 7.96E -04 2.70% Pons cstL 6.65E -04 5.27E -04 4.93E -04 6.62E -04 6.19E -04 5.93E -04 11.90% cstR 6.88E -04 5.85E -04 5.67E -04 6.60E -04 5.93E -04 6.18E -04 7.60% Medulla cstL 9.10E -04 1.06E -03 1.12E -03 8.05E -04 8.99E -04 9.58E -04 12.00% cstR 1.02E -03 1.03E -03 1.02E -03 8.53E -04 1.01E -03 9.87E -04 6.80% Fractional Anisotropy (FA) MidBrain peduncleL 6.93E -01 6.18E -01 6.65E -01 6.02E -01 6.68E -01 6.49E -01 5.20% peduncleR 6.80E -01 6.39E -01 6.37E -01 5.60E -01 6.41E -01 6.31E -01 6.20% periaqueductL 2.58E -01 2.29E -01 2.64E -01 2.11E -01 2.31E -01 2.39E -01 8.20% periaqueductR 2.21E -01 1.86E -01 2.36E -01 2.26E -01 2.15E -01 2.17E -01 7.80% Pons cstL 4.72E -01 5.50E -01 6.36E -01 5.83E -01 5.27E -01 5.53E -01 9.90% cstR 4.44E -01 5.34E -01 5.48E -01 5.14E -01 5.58E -01 5.19E -01 7.80% Medulla cstL 4.38E -01 4.40E -01 4.12E -01 4.70E -01 5.07E -01 4.54E -01 7.20% cstR 3.78E -01 4.19E -01 3.99E -01 4.43E -01 4.03E -01 4.08E -01 5.20% the ROIs in the midbrain, pons, and medulla (Fig. 2). Average diffusivity is a measure of the mean amount of diffusion in a magnetic resonance voxel.18 It only provides indirect information on the integrity of axon and extra axonal matrix in the brain tissues19. When axons are demyelinated, the loss of cell and matrix integrity is reflected as a higher AD value, because the presence of cell membranes and matrix molecules slows the rate of diffusion, represented by a lower AD value. This diffusion parameter is calculated through use of the following equation ( Eq. 2)16 = F ractional anisotropy (FA) is a diffusion parameter indicating the directionality of the water diffusion. Fractional anisotropy values were calculated using the equation below (Eq. 3).16 = 1 2 ( ) + ( ) + ( ) + + (Eq. 3) A low FA value designates that the water diffusion is directionally independent (i.e., it is random and unrestricted). Relatively high FA values classify the water diffusion as directionally dependent on restricted diffusion along cell membranes (e.g., unidirectional water diffusion through a nerve fiber).20 Degradation of the brain stem nerve structures will increase the isotropy of water diffusion, resulting in a lower FA value. In this manner, FA val ues can indicate whether a nerve bundle is intact (high FA value) or compromised (low FA value).21 (Eq. 2)
SHEL BY GOICOCHEA, LUIS COLONPEREZ, DAVID B. FITZGERLAD, THOMAS MARECI University of Florida | Journal of Undergraduate Research | Volume 15, Issue 3 | S ummer 2014 4 DISCUSSION Diffusion weighted images improved the brain stem structures contrast, which allowed for appropriate segmentation and ROI delineation; the ROIs were easily identified in the AD and FA maps. High FA values can be obser ved in white matter structures (peduncle and CST) and low values are observed in gray matter regions (periaqueduct), which agrees with expectation for the tissue structure. The relative standard deviation (RSD) of the AD and FA values is moderately large, but can be significantly decreased by collecting a larger population of healthy brain stem data. This would also create a more reliable and robust brain stem diffusion value database. In future work, brain stem images will be collected from veterans clinically diagnosed with disrupted sleep and mild traumatic brain injury. The calculated diffusion values will be compared to the healthy subject database to quantify the effects of TBI on the brain stem. We suspect that the diseased population will have increa sed AD values and decreased FA values due to damage of the brain stem tracts as a result of blast or blunt trauma. ACKNOWLEDGEMENTS Funding was provided to the author by the University Scholars Program, University of Florida (UF). The data was acquired at the Advanced Magnetic Resonance Imaging & Spectroscopy (AMRIS) Facility at the McKnight Brain Institute of the UF. Additional funding was provided by the State of Florida Brain and Spinal Cord Injury Research Trust Fund, VA RR+D CDA 2 Award B6698W*, and N IH RO1 NS063360#. REFERENCES 1. Faul M, Xu L, Wald MM, Coronado VG. Traumatic brain injury in the United States: emergency department visits, hospitalizations, and Figure 2. Left panel: AD and FA maps of in vivo brain stem acquisitions with ROIs drawn following selection criteria detailed in the methods section. For the midbrain ROIs (A) the cerebral crus (top) and cerebral aqueduct (bottom) were drawn in an FA and AD map, respectively. FA map were used in the R OI drawings of (B) corticospinal tract (cst) and (C) pyramid of medulla. Right panel: Brain stem histological slices from Duvernoys Atlas15. (D) 1. Cerebral crus, 9. Cerebral aqueduct (E) 2. CST (F) 1. Pyramid of medulla with CST
University of Florida | Journal of Undergraduate Research | Volume 15, Issue 3 2 | Summer 2014 5 deaths in 2002 -2006. Atlanta, GA: U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, National Center for Injury Prevention and Control; 2010: 5-9. 2. Coronado VG, Xu L, Basavaraju SV, McGuire LC, Wald MM, Faul MD, Guzman BR, Hemphill JD. Surveillance for traumatic brain injury -related deaths United States, 1 997 -2007. Surveillance Summaries. 2011; 60(5):1-32. 3. Orff HJ, Ayalon L, Drummond SPA. Traumatic brain injury and sleep disturbance: a review of current research. J Head Trauma Rehabil 2009; 24(3): 155165. 4. Ayalon L, Borodkin K, Dishon L, Kanety H, Da gan Y. Circadian rhythm sleep disorders following mild traumatic brain injury. Neurology 2007; 68(14): 1136-1140. 5. Ouellet, MC, Morin CM. Subjective and objective measures of insomnia in the context of traumatic brain injury: a preliminary study. Sleep Medicine. 2006; 7: 486497. 6. Williams BR, Lazic SE, and Ogilvie RD. Polysomnographic and quantitative EEG analysis of subjects with long -term insomnia complaints associated with mild traumatic brain injury. Clin Neurophysiol 2008; 119(2): 429438. 7. Sh ekleton JA, Parcell DL, Redman JR, Phipps -Nelson J, Ponsford JL, Rajaratnam SM. Sleep disturbance and melatonin levels following traumatic brain injury. Neurology 2010; 74(21): 17321780. 8. Berry, RB. Neurobiology of Sleep. In: Goolsby J, Pritchard J, eds. Fundamentals of Sleep Medicine Philadelphia, PA: Elsevier; 2012: 91100. 9. Blumenfeld H. Brainstem III: internal structures and vascular supply. Neuroanatomy Through Clinical Cases 2nd edition. Sunderland, MA: Sinauer; 2002: 598 -600. 10. Le Bihan D, Mangin JF, Poupon C, Clark CA, Pappata S, Molko N, Chabriat H. Diffusion tensor imaging: concepts and applications. J Magn Reson Imaging. 2001; 13: 534 -546. 11. Basser PJ, Mattiello J, Le Bihan D. Estimation of the effective self diffusion tensor from the NMR spin-echo. J Magn Reson B 1994; 103: 247 254. 12. Le Bihan D, Turner R, Pekar J, Moonen CT. Diffusion and perfusion imaging by gradient sensitization: design, strategy and significance. J Magn Reson Imaging. 1991; 1: 7 8. 13. Smith SM, Jenkinson M, Woolrich MW, Beckmann CF, Behrens TE, Johansen-Berg H, Bannister PR, De Luca M, Drobnjak I, Flitney DE, Niazy RK, Saunders J, Vickers J, Zhang Y, De Stefano N, Brady JM, Matthews PM. Advances in functional and structural MR image analysis and implementation as FSL. Neuroimage 2004; 23 (suppl 1): S208219. 14. Ford A, Triplett W. Personal oral communication concerning ROI placement selection at the McKnight Brain Institute, University of Florida. January 2012. 15. Naidich TP, Duvernoy HM, Delman BN, Sorensen GA, Kollias SS, Haacke EM. Duvernoy's Atlas of the Human Brain Stem and Cerebellum: High -Field MRI, Surface Anatomy, Internal Structure, Vascularization and 3D Sectional Anatomy Austria: Sprin ger; 2009: 26888 16. Stejskal EO, Tanner JE. Spin diffusion measurements: spin echoes in the presence of a time-dependent field gradient. J Chem Phys 1965; 42: 288 292. 17. Basser PJ, Mattiello J, Le Bihan D. MR diffusion tensor spectroscopy and imaging. Biophys J 1994; 66: 259 267. 18. Farrell JAD, Landman BA, Jones CK, Smith SA, Prince JL, van Zijl PCM, Mori S. Effects of signal to -noise ratio on the accuracy and reproducibility of diffusion tensor imaging derived fractional anisotropy, mean diffusivit y, and principal eigenvector measurements at 1.5T. J Magn Reson Imaging.2007; 26: 756-767. 19. Cercignani M, Inglese M, Pagani E, Comi G, Filippi M. Mean diffusivity and fractional anisotropy histograms of patients with multiple sclerosis. Am J Neuroradiol 2001; 2: 952-958. 20. Basser PJ, Pierpaoli C. Microstructural and physiological features of tissues elucidated by quantitative -diffusion-tensor MRI. J Magn Reson Imaging. Series B. 1996; 111: 209-219. 21. Parekh MB, Ca rney PR, Sepulveda H, Norman W, King M, Mareci T. early MR diffusion and relaxation changes in the parahippocampal gyrus precede the onset of spontaneous seizures in an animal model of chronic limbic epilepsy. Exp Neurol 2010; 224: 258 -270.