<%BANNER%>

Nonfluent Aphasia

Permanent Link: http://ufdc.ufl.edu/UFE0024429/00001

Material Information

Title: Nonfluent Aphasia The Relationship Between Degree of Left-Hemisphere Lesion, Homologous Brain Activity, and Performance
Physical Description: 1 online resource (49 p.)
Language: english
Creator: Zlatar, Zvinka
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: aphasia, fmri
Clinical and Health Psychology -- Dissertations, Academic -- UF
Genre: Psychology thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Little is known about the predictive value of degree of lesion on the reorganization of language function and recovery in aphasia. This study investigated whether degree of lesion in left-perisylvian regions of interest (ROIs) was associated with the amount of right-hemisphere homologous functional activity during word generation, and performance on the Boston Naming Test (BNT) and Western Aphasia Battery-Aphasia Quotient measures (WAB-AQ). Nineteen chronic, nonfluent aphasia patients received structural and functional MRI prior to an intention-based language treatment aiming to re-lateralize language functions to the right hemisphere. During imaging, patients performed an event-related category member generation task. Two independent raters scored the degree of lesion in the left hemisphere on a scale of 0 to 5 and correlated the ratings with the volume of functional activity in the right-hemisphere homologues and the BNT and WAB scores prior to treatment. There was no relationship between degree of lesion in the left hemisphere and the amount of functional activity in the right-hemisphere homologous structures. However, patients with a higher degree of lesion showed poorer performance on measures of fluency and repetition. The authors concluded that reorganization of language functions to the right-hemisphere homologous structures shown in the literature may not be a universally applicable concept and that degree of lesion is marginally related to spontaneous speech and repetition abilities in this sample.
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 Zvinka Zlatar.
Thesis: Thesis (M.S.)--University of Florida, 2009.
Local: Adviser: Crosson, Bruce A.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024429:00001

Permanent Link: http://ufdc.ufl.edu/UFE0024429/00001

Material Information

Title: Nonfluent Aphasia The Relationship Between Degree of Left-Hemisphere Lesion, Homologous Brain Activity, and Performance
Physical Description: 1 online resource (49 p.)
Language: english
Creator: Zlatar, Zvinka
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2009

Subjects

Subjects / Keywords: aphasia, fmri
Clinical and Health Psychology -- Dissertations, Academic -- UF
Genre: Psychology thesis, M.S.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Little is known about the predictive value of degree of lesion on the reorganization of language function and recovery in aphasia. This study investigated whether degree of lesion in left-perisylvian regions of interest (ROIs) was associated with the amount of right-hemisphere homologous functional activity during word generation, and performance on the Boston Naming Test (BNT) and Western Aphasia Battery-Aphasia Quotient measures (WAB-AQ). Nineteen chronic, nonfluent aphasia patients received structural and functional MRI prior to an intention-based language treatment aiming to re-lateralize language functions to the right hemisphere. During imaging, patients performed an event-related category member generation task. Two independent raters scored the degree of lesion in the left hemisphere on a scale of 0 to 5 and correlated the ratings with the volume of functional activity in the right-hemisphere homologues and the BNT and WAB scores prior to treatment. There was no relationship between degree of lesion in the left hemisphere and the amount of functional activity in the right-hemisphere homologous structures. However, patients with a higher degree of lesion showed poorer performance on measures of fluency and repetition. The authors concluded that reorganization of language functions to the right-hemisphere homologous structures shown in the literature may not be a universally applicable concept and that degree of lesion is marginally related to spontaneous speech and repetition abilities in this sample.
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 Zvinka Zlatar.
Thesis: Thesis (M.S.)--University of Florida, 2009.
Local: Adviser: Crosson, Bruce A.

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2009
System ID: UFE0024429:00001


This item has the following downloads:


Full Text

PAGE 1

NONFLUE NT APHASIA: THE RELATIONSHIP BETWEEN DEGREE OF LEFTHEMISPHERE LESION, HOMOLOGOUS BR AIN ACTIVITY, AND PERFORMANCE By ZVINKA ZOE ZLATAR A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2009 1

PAGE 2

2009 Zvinka Zoe Zlatar 2

PAGE 3

To m y parents and wonderful family 3

PAGE 4

ACKNOWL EDGMENTS I thank my mentor, Dr. Bruce Crosson, for his help and guidance thr oughout this project. I must also express my gratitude to Ilana Levy w ho was an essential contri butor to this study and spent countless hours rating brain lesions with me. Likewise, I want to thank the following members of the Brain Imaging Rehabilitati on and Cognition laboratory for their support: Matthew Cohen, Jonathan Trinastic Keith McGregor, and Michelle Benjamin. This research was funded by grants #P50 DC03888 and #R01 DC007387 from the National In stitute on Deafness and Other Communication Disorders and by Center of Excellence grant # F2182C and Research Career Scientist Award # B3470S from the Depa rtment of Veterans Affairs Rehabilitation Research and Development Service. 4

PAGE 5

TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........7 ABSTRACT.....................................................................................................................................8 CHAPTER 1 INTRODUCTION................................................................................................................. .10 Reorganization of Language Functions in Aphasia................................................................10 Degree of Lesion and Language Performance in Aphasia.....................................................14 Aim 1......................................................................................................................................16 Aim 2......................................................................................................................................17 2 METHODS...................................................................................................................... .......18 Subjects...................................................................................................................................18 Procedure................................................................................................................................19 Testing Session................................................................................................................19 Imaging Task: Category Member Generation.................................................................20 Image Acquisition...........................................................................................................21 Image Analyses...............................................................................................................22 Lesion Analysis...............................................................................................................24 Statistical Analyses........................................................................................................... ......26 Analysis 1........................................................................................................................26 Analysis 2........................................................................................................................27 3 RESULTS...................................................................................................................... .........29 Results for Aim 1.............................................................................................................. ......29 Results for Aim 2.............................................................................................................. ......31 4 DISCUSSION................................................................................................................... ......36 Review of Findings.................................................................................................................36 Review of Aim 1.............................................................................................................36 Review of Aim 2.............................................................................................................38 Study Implications..................................................................................................................39 Study Limitations.............................................................................................................. ......40 Future Directions....................................................................................................................41 5

PAGE 6

APPENDIX A DEGREE OF LESION RATING SCALE.............................................................................42 B DEGREE OF LESION SCORING SHEET...........................................................................43 C ROI DEFINITIONS.............................................................................................................. ..44 LIST OF REFERENCES...............................................................................................................45 BIOGRAPHICAL SKETCH.........................................................................................................49 6

PAGE 7

LIST OF TABLES Table page 2-1 Subject demographics (N=19)..........................................................................................28 2-2 ROI names and abbreviations...........................................................................................28 3-1 Correlations between Anterior and Po sterior ROI lesion scores and homologous activity (N = 10).............................................................................................................. ...32 3-2 Number of right-hemisphe re active voxels (N = 10)........................................................32 3-3 Left-hemisphere lesion ratings (N= 19)............................................................................32 3-4 Individual lesion ratings....................................................................................................33 3-5 Behavioral language measures (N = 19)...........................................................................33 3-6 Correlations between ROIs, homologous activity, and language performance (N = 10)......................................................................................................................................34 3-7 Correlations between lesion scores and language performance (N = 19).........................35 7

PAGE 8

Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science NONFLUENT APHASIA: THE RELATIONSHIP BETWEEN DEGREE OF LEFTHEMISPHERE LESION, HOMOLOGOUS BRAI N ACTIVITY, AND PERFORMANCE By Zvinka Zoe Zlatar May 2009 Chair: Bruce Crosson Major: Psychology Little is known about the pr edictive value of degree of le sion on the reorganization of language function and recovery in aphasia. This study investigat ed whether degree of lesion in left-perisylvian regions of inte rest (ROIs) was associated with the amount of right-hemisphere homologous functional activity during word gene ration, and performance on the Boston Naming Test (BNT) and Western Aphasia Battery-Aphasi a Quotient measures (WAB-AQ). Nineteen chronic, nonfluent aphasia patients received stru ctural and functional MRI prior to an intentionbased language treatment aiming to re-laterali ze language functions to the right hemisphere. During imaging, patients performed an eventrelated category member generation task. Two independent raters scored the degree of lesion in the left hemisphere on a scale of 0 to 5 and correlated the ratings with the volume of func tional activity in the right-hemisphere homologues and the BNT and WAB scores prior to treatment There was no relations hip between degree of lesion in the left hemisphere and the amount of functional activity in the right-hemisphere homologous structures. However, patients w ith a higher degree of lesion showed poorer performance on measures of fluency and repetitio n. The authors concluded that reorganization of language functions to the right-hemisphere hom ologous structures shown in the literature may 8

PAGE 9

9 not be a universally applicable concept and that degree of le sion is marginally related to spontaneous speech and repetiti on abilities in this sample.

PAGE 10

CHAP TER 1 INTRODUCTION Reorganization of Language Functions in Aphasia Even before the advent of functional imaging, scientists began to que stion the role of the right hemisphere in recovery from aphasia. In 1877, Barlow reported the case of an aphasic patient who recovered language functions followi ng a lesion in Brocas area. The same patient later lost language abilities s ubsequent to a right-hemisphere stroke. Based on this case, Barlow suggested that right-hemisphere mechanisms were involved in the recovery of language functions in aphasia (Barlow, 1877). This claim has been supported by ot her studies indicating that language functions in aphasic patients beco me impaired when the right hemisphere is anesthetized during WADA tests (Kinsborne, 1971). Similarly, Basso and colleagues have reported cases in which neuropsycho logical test performance decrea sed in aphasic patients after subsequent right-hemisphere str oke (Basso, Gardelli, Grassi, & Mariotti, 1989). Hence the right hemisphere seems to play an important role in the recovery of langua ge functions in some aphasia patients. Current evidence regarding left and right-hemis phere contributions to language function in aphasia remains controversial. A number of studi es have indicated that recovery of language function in aphasia is primarily involved with right-hemisphere activity (Abo et al., 2004; Crosson et al., 2005; Gold & Kertesz, 2000; Weille r et al., 1995), while others emphasize the importance of dominant perilesion al activity in recovery (Breier et al., 2004; Cornelissen et al., 2003; Duffau, Bauchet, Leheric y, & Capelle, 2001; Leger et al., 2002; Miura et al., 1999; Seghier et al., 2001; Warburton, Price, Swinbur n, & Wise, 1999). For example, using fMRI, one study investigated how brain activ ity during meaningful and mean ingless stories changed with the patients' auditory sentence comprehension ski lls. The authors found that irrespective of lesion 10

PAGE 11

site, performance on tests of auditory sentence comprehension was positively correlated with activation in the right lateral superior te m poral region (Crinion & Price, 2005). Other evidence supporting the involvement of ri ght-hemisphere activity in recovery from aphasia has suggested that when a shift of language functi ons occurs, it is mainly reorganized to right-hemisphere structures homologous to damaged left-hemisphere areas (Lazar et al., 2000). For example, Weiller and colleagues demonstrated that patients who had recovered from Wernickes aphasia, after damage to Wernickes area, showed activity in the right-hemisphere homologue of Wernickes area (Weiller et al ., 1995). Similarly, Blank and colleagues showed that aphasic patients with lesions in the left pars opercularis (posterior part of Brocas area) had more pronounced activity in the right pars opercularis during a language production task than aphasic patients without left pars opercularis le sions and neurologically h ealthy control subjects (Blank, Bird, Turkheimer, & Wise, 2003). Just as some investigators advocate that recovery from aphasia is enhanced when the right hemisphere takes over language functions from da maged left-hemisphere areas, others suggest that optimal recovery in aphasia is acco mpanied by reorganization to left-hemisphere perilesional areas (Cao, Viki ngstad, George, Johnson, & Welch, 1999; Heiss, Kessler, Thiel, Ghaemi, & Karbe, 1999; Karbe et al., 1998). Hei ss and colleagues found that aphasia recovery, measured by a test of language comprehensi on, was correlated with the re-activation of perilesional left-hemisphere struct ures, and that poor recovery was associated with activity in certain areas of the right hemisphere (Heiss et al., 1997). Similarly, one study used repetitive transcranial magnetic stimulation (rTMS) to inactivate the right pars triangularis (analog to the anterior part of Brocas area) as a treatment for aphasia. They demonstrated that all of the patients who received the rTMS treatment s howed improved language performance two months 11

PAGE 12

af ter treatment, indicating that activity in right pars triangularis may have hindered recovery for patients included in the study (Mar tin et al., 2004). However, the same author also found that rTMS of the right pars opercularis (analog to the posterior part of Broca s area) decreased picture naming accuracy and increased response latencie s (Naeser et al., 2002). Th ese findings indicate that certain right-hemisphere structures may in terfere with language functioning while others may actually facilitate it. The previously reported findings raise the question of degree of lesi on and its role in the re-organization of language functi ons in aphasia. It may be possible that aphasia patients with smaller lesions do not need to re -lateralize language functions to the right hemisphere, due to the preservation of key structures of the language network in the dominant hemisphere. However, patients with more severe left-hemisphere lesi ons may reorganize certai n language functions to the non-dominant hemisphere because the dominant hemisphere may no longer be a viable substrate for recovery (Crosson et al., 2007). Pa rallel to this point of view, Heiss and Thiel proposed a hierarchy of recovery based on degree of lesion, where best recovery of function can be associated with the reactiv ation of the dominant hemisphere, which is only possible after small lesions that affect language areas of minor importance to the network. The next step in the hierarchy is the involvement of an intrahemisphe ric network concerning s econdary centers of the dominant network, which may lead to satisfact ory improvement. The authors finally suggested that a less efficient interhemispheric compen sation mechanism involving the contralateral homologous areas is related to le ss efficient improvement and is dependent on the severity of the damage to network components that reduce transca llosal inhibition (Heiss & Thiel, 2006). If this is the case, aphasia patients with higher degree of lesion in speci fic left-hemisphere structures, 12

PAGE 13

should exhibit m ore pronounced activity in the right hemisphere, specifically in the homologous structures, than those wi th less severe lesions. The work of Blank and colleagues (2003) s uggested that right-hemisphere areas are recruited during language production only when their left-hemisphere counterparts are damaged. They used positron emission tomography (PET) to study the language production of 14 former nonfluent aphasia patients who subs equently recovered the ability to spea k spontaneously in sentences. Patients performed an everyday pr opositional speech task during imaging. The authors found that aphasia patients with lesions to the left pars ope rcularis showed more pronounced activity, during a narrative language task, than healt hy controls and aphasic patients with little to no lesion in the left pars operc ularis. Since the patients had already recovered language functions at the time of assessment, th e authors suggested that the increased activity observed in the right pars opercularis was related to recovery fr om nonfluent aphasia. A major weakness in Blanks work is that they only test ed their hypothesis for pa rs opercularis and did not take into account the degree of lesion and its potential im pact on the amount of activity observed in the homologous structures. It is importa nt to replicate and extend the work of Blank and colleagues to include more areas of the langu age network and determine if degree of lesion in specific regions is associated with the amount of homologous act ivity in the right hemisphere. This study examined the relationship between de gree of lesion in the left hemisphere and the amount of homologous right-hemisphere activity during word ge neration in nonfluent aphasia patients. It was hypothesize d that patients with higher degr ee of lesion in anterior and posterior perisylvian regions of the language network would show more pronounced activity in the right-hemisphere homologous structures due to the diminished capacity of the dominant 13

PAGE 14

hem isphere to provide the necessary substrates for recovery, thereby enha ncing re-lateralization of language functions to the right-h emisphere homologous structures. Degree of Lesion and Language Performance in Aphasia Little research has examined the relationship between degree of left-hemisphere lesion and language performance in aphasia. The extant l iterature regarding lesion characteristics and aphasia recovery appears perplexi ng due to considerable variation with regard to the type of aphasia patients included in the samples, the em ployment of inconsistent inclusion criteria, and the vast array of behavioral langua ge measures used to assess rec overy. Most of the research in this area has focused on the relationship between le sion characteristics and aphasia type (Kreisler et al., 2000), or the behavioral predictors of recovery in aphasia, such as measures of language comprehension predicting greater treatment su ccess (Naeser et al., 1998). Unfortunately, the literature has neglected to consistently addre ss the relationship between degree of lesion in specific areas of the language network and perf ormance on different language measures during recovery in nonfluent aphasia. One study found that lesions in the posteriorsuperior-temporal and supramarginal regions were associated with poor auditory comprehens ion in a sample of 39 left-hemisphere stroke patients (Selnes, Knopman, Niccum, Rubens, & La rson, 1983). On the contra ry, a different study found that lesions to Broca's and Wernicke's areas were not significantl y related to language comprehension (Dronkers, Wilkins, Van Valin, Redfern, & Jaeger, 2004). The authors used voxel-based lesion symptom mapping (VLSM) to investigate the relationship between lefthemisphere damaged areas of 64 chronic stroke patients and performance on a measure of language comprehension. They found that th e following left-hemisphere areas affected performance: the posterior middle temporal gyr us and underlying white matter, the anterior 14

PAGE 15

superior tem poral gyrus, the supe rior temporal sulcus and angula r gyrus, mid-frontal cortex in Brodmann's area 46, and Broadmanns area 47 of the inferior frontal gyrus. A subsequent study used VLSM on 50 aphasic patients to inves tigate the lesion correlates of conversational speech deficits. The authors found that damage to the anterior insula predicted low grammatical complexity and amount of speech produced. Additionally, the inferior frontal gyrus, sensorimotor and anterior temporal areas were also associ ated with lower scores on both grammatical complexity and amount of speech produced (Borovsky, Saygin, Bates, & Dronkers, 2007). Similarly, it has been report ed that subjects with overa ll larger lesions showed less recovery of naming abilities whil e subjects with lesions in Wern ickes area and the inferior parietal cortex demonstrated the most severe naming impairment (Knopman, Selnes, Niccum, & Rubens, 1984). In sum, the majority of studies have suggest ed that the intactness of posterior regions, especially Wernickes area, is important for language recovery in aphasia. It is important to bear in mind that the current findings are inconsistent in terms of type a nd severity of aphasic disorder, procedures used to measure lesion size, brain areas included as re gions of interest, and language measures used to assess improvement. Furthermore, individu al cases have been reported which contradict most group findings (Basso & Farabola, 1997). Due to the inconsistent findings in the literature, more research is need ed to better understand the relationship between degree of lesion and behavioral language measur es associated with recovery of language functions in non-fluent aphasia. The main purpose of this study was to expand th e existent literature regarding the role of degree of lesion in the amount of functional ho mologous activity and language performance in nonfluent aphasia. One of the main goals was to attempt to replicate and extend the findings of 15

PAGE 16

Blank and colleagues (B lank et al., 2003) to examin e the relationship between degree of lesion in the left hemisphere and the amount of homo logous right-hemisphere activity during word generation in chronic, nonfluent aphasia patient s. Degree of lesion and homologous activity were measured in two broad regions of interest (ROIs): anterior (A ) and posterior (P) perisylvian regions. The A ROI consisted of pars opercularis and pars triangularis, which together form Brocas area. The P ROI was comprised of the supramarginal gyrus, the angular gyrus, and the posterior third of the superior a nd inferior temporal gyri, which was labeled Wernickes +. These areas were selected because they all serve diffe rent aspects of language function and have been associated with recovery in aphasia (Blank et al., 2003; Knopman et al ., 1984; Martin et al., 2004; Selnes et al., 1983). For a detailed definition of the anatomical boundaries of each ROI, see Appendix C. Another goal of this study was to examine the relationship between degree of lesion and language performance in nonfluent aphasia. As stat ed in the introduction, the literature remains inconclusive regarding th e role of degree of lesion in la nguage functions in nonfluent aphasia due to the variability of aphasi a types included in the samples and the measures used to assess recovery. This study examined the relationship between degree of lesion in the A and P ROIs and performance in nonfluent aphasia using th e Boston Naming Test [BNT] (Kaplan, Goodglass, & Weintraub, 1983) and the Western Aphasia Ba ttery-Aphasia Quotient measures [WAB-AQ] (Kertesz, 1982). Aim 1 To examine the relationship between left-hemis phere degree of lesion in the A and P ROIs and amount of right-hemisphere homologous acti vity during word generation in nonfluent aphasia. 16

PAGE 17

17 Hypothesis 1: It was hypothesized that patients with higher degree of lesion in the A and P regions would show more pronounced activity in the right-hemisphere homologous structures of the same ROIs. More pronounced homologous activity may be due to the diminished capacity of the dominant hemisphere to provide the necessary substrates for recove ry, thereby encouraging re-lateralization of language f unctions to the right-hemisphere homologous structures (Heiss & Thiel, 2006). Aim 2 To examine the relationship be tween degree of lesion in the left-hemisphere A and P ROIs and language performance in nonfluent apha sia using the BNT and WAB-AQ measures. Hypothesis 2: Due to the inconsistent literature regarding the relationship between degree of lesion and language performance in aphasia, exploratory analyses were carried out to investigate the role of degree of lesion in aphasia recovery using the BNT and WAB-AQ measures. It was hypothesized th at higher degree of lesion in the A and P ROIs would be associated with lower scores on the BNT and WAB-AQ measures [spontaneous speech, comprehension, repetition, and naming] (Goldenbe rg & Spatt, 1994; Kertesz et al., 1993; M. Naeser et al., 1990).

PAGE 18

CHAP TER 2 METHODS Subjects A subset of subjects from two larger apha sia treatment studies c onsisting of 19 patients with nonfluent aphasia was included in this samp le. These studies (1 and 2) consisted of an intention-based language treatment for chronic nonfluent aphasia patients and aimed to relateralize language functions to the right medio-frontal hemisphere by asking subjects to initiate word production trials with a complex, non-symbo lic left-hand gesture. Both samples did not differ significantly from each other on any variable, with exception of the degree of lesion in the P ROI, where study 1 had a mean posterior sc ore of 3.38 (SD = 1.75) while study 2 had a mean posterior score of 6.55 (SD = 4.34) [ t (13.97) = -2.19, p < .05]. The current study included data from pre treatment fMRI and language measures only, and therefore did not investigate changes in brain activity or language improvement from pre to post treatment. Of the 19 patients, only 10 had usable fMRI im ages; therefore, data from 10 subjects was used to examine the relationship between de gree of lesion and right-hemisphere homologous activity, while data from all 19 subjects was us ed to assess the relationship between degree of lesion and language performance. See Table 2-1 for subject demographics. All subjects were recruited from the Brain Rehabilitation Research Center of the Malcom Randall VA Medical Center in Ga inesville, FL, the Shands Health Care System, Brooks Rehabilitation Hospital in Jacksonville, FL, and the Gainesville and Jacksonville communities at large. All subjects had documented left-hemisphere lesions on clinical MRI or CT scans due to ischemic or hemorrhagic stroke. Subjects had no history of right-hemis phere stroke, learning disabilities, neurological conditions, or chronic substance abus e and were premorbidly righthanded and native English speakers. All subjects gave informed consent according to the 18

PAGE 19

guidelines stipulated by the Intern al Review Board (IRB) of the University of Florida Health Science Center. Inclusion criteria: Non-fluency was defined by hesitatio ns due to word-finding difficulty on a narrative speech sample elicited from the Boston Diagnostic Aphasia Examinations Cookie Theft Picture (Goodglass & Kaplan, 1983). S ubjects also had minimally impaired comprehension at the single-word level and eviden ce of posterior inferior frontal lesion on MRI scans. All subjects were 6 or more months pos t onset of their most recent stroke and had no contraindications for MRI scanning, such as ca rdiac pacemaker, ferrous metal implants, or claustrophobia. Word-finding difficulty was moderate to severe as indicated by scores of less than 48 and more than 3, out of 60 naming items, on the Bo ston Naming Test (Kaplan et al., 1983). To ensure that patients understood st udy instructions during the fMRI session, study 2 included only those patients with a score within two standard deviations from the ag e-appropriate mean for normal subjects on the Peabody Picture Vocabul ary Test IV (Dunn & Dunn, 2007). See Table 21 for subject demographics. Procedure Structural and functional MRI and behavioral testing data (BNT and WAB) were obtained during the subjects pr e-treatment scanning and testing se ssions, respectively, which were usually less than one week apart. Even though a total of 19 subjects participated in fMRI and testing sessions, only 10 subjects had usable fMRI data. Testing Session A variety of behavioral measur es were acquired as part of the protocol for the larger studies, however only those measures relevant to this particular study will be discussed. 19

PAGE 20

All subjects were adm inistered the BNT and the WAB-AQ subtests. The BNT is a confrontation naming test consis ting of 60 items (Kaplan et al., 1983). Subjects attempted to name all 60 line drawings. For the final analys es the percent correct, out of 60 items, was employed. Subjects were also administered th e Western Aphasia Battery (WAB) subtests necessary to calculate the Aphasia Quotient ( AQ). This portion of the test was developed to assess the main clinical aspects of language function: content, fluency, auditory comprehension, repetition and naming. The aphasia quotient is an overall measure of aphasia severity which uses the oral portion of the WABs language a ssessment (Kertesz, 1982). Subjects WAB AQ, spontaneous speech, comprehension, repetition, a nd naming raw scores were entered into the analyses. Imaging Task: Category Member Generation The fMRI task used to examine the level of activity in right-hemis phere homologues of the A and P ROIs was an event-related overt word generation task. For study 2, participants saw and heard a category (e.g., birds) and responded by saying aloud a single category member (e. g., robin). There were 6 runs of 10 category exemplars each for a total of 60 category-member generation trials. Each active trial lasted 13.6 seconds (8 TRs). The inter-trial interv als alternated between 13.6, 15.3, and 17 seconds in a pseudo random fashion, which allows adequate time for the hemodynamic responses of aphasic patients to retu rn to baseline before the subsequent trial is initiated. Each run lasted 5 minutes, 16.2 sec onds for a total fMRI sequence of 31.62 minutes. For study 1, the category member generation task c onsisted of 5 runs with 9 category exemplars each, for a total of 45 trials. Inter-trial in tervals varied from 21.6, 23.2, 24.9, and 26.6 seconds and were randomized over the course of the ru ns. The categories for study 1 were presented auditorily only rather than audio-visually. 20

PAGE 21

The auditory stim uli were presented via ea rphones of a Commander XG Audio System (Resonance Technology) built specifi cally for use in MRI environments. The visual stimuli, for study 2 only, were presented via a computer mo nitor mounted over the head coil (Eloquence systems). The category-member generation task was selected because previous studies in our laboratory have found that, compared to other la nguage tasks; it optimizes activity in languagerelated areas including the RO Is selected in this study. An Eloquence System (INVIVO) was used to prog ram the experiment. It is equipped with E-Prime software to program experiments a nd record data. The output of the Eloquence computer was connected to the MRI Audio System for stimulus presenta tion. Subjects verbal responses on word-generation tasks were record ed by connecting a PhoneOr dual channel fiber optic, noise-canceling microphone to a separa te Dell laptop computer using Audition 1.5 (Adobe) software for subjects responses to be scored off-line. Similarly, subjects responses were recorded with Cool Edit 2000 software to obtain response times off-line. Image Acquisition Since data were obtained from studies 1 and 2, structural and functional images were acquired using three different 3 Tesla scanners. Im ages for the subjects in study 1 (N = 8) were acquired on a GE 3T Signa LX scanner with a qu adrature radio frequency coil, or on a Siemens Allegra 3T scanner. Thirty-two contiguous sagi ttal slices covering whol e brain were acquired using a 1-shot spiral sequence (TR = 1660 ms; TE = 18 ms; 70 flip angle; 64x64 matrix; FOV = 200 mm; 4 mm in thickness). An additional 6 images (9.96 seconds) were added to the beginning of each functional run to allow for homogeneity of the MR signal. A structural T1-weighted spoiled gradient-recalled acquisition in the steady state (GRA SS) sequence was obtained for localization purposes and to determine the degree of lesion in the left-hemisphere ROIs (TR = 23 21

PAGE 22

m s; TE = 6 ms; 25 flip angle; 256x192 matrix ; 1.3 mm in thickness; 124 slices; FOV = 240 mm). Foam padding was used to limit head motion during scanning. All images for study 2 (N = 11) were acquired on a Philips 3 Tesla whole body scanner at the McKnight Brain Institute of the University of Florida. For functiona l MRI, the whole brain was imaged in 1.7 seconds using a gradient echo planar sequence with 36 x 4 mm thick sagittal slices (TR = 1700 ms; TE = 30 ms; FA = 70 degrees ; FOV = 24 cm, matrix = 64 x 64). Prior to collecting functional images, highresolution T1-weighted structural images were acquired for 160 x 1 mm thick sagittal slices using TFE acquisition (TE = 3.7 ms; TR = 8.1 ms; FOV = 24 cm; FA = 8 degrees; matrix size = 240 x 240). The T1-weigthed images were used to assess the degree of lesion in the left-hemisphere ROIs using a rating scale from 0-5, while the functional images were used to obtain the number of ac tive voxels on each of the right-hemisphere homologous ROIs. Image Analyses Images were analyzed on a Linux worksta tion (Dell) using Analysis of Functional Neuroimages (AFNI) software (Cox, 1996). To redu ce effects of head motion, the 3-dimensional volume registration option in AFNI was used to spatially register images from each run to the first image of the first functional run, because its acquisition immediatel y follows acquisition of T1-weigthed anatomic images onto which functi onal images are overlaid. All voxels in which the standard deviation (SD) of acquired time se ries exceeded 8% of the mean signal intensity was set to zero to minimize large vessel effects an d other artifacts. Individual imaging runs were orthogonalized for linear trends and then concaten ated into a single time series. Due to residual motion artifacts produced by overt language production in the sca nner, functional images were detrended for motion artifacts using a refinement of the procedure developed by Gopinath and 22

PAGE 23

colleagues (Gopinath et al., in press). Analyses included a ll trials in which a participant attempted a spoken response. Hemodynamic responses (HDRs) for word-generation trials were modeled as a single HDR using the deconvolution program in AFNI. Deconvolution makes no a priori assumptions about the shape of the HDR, but rather derives it empirically. In deconvolution, an estimate of the HDR is deconvolved from the acquired time seri es in each voxel; then, the estimated HDR is convolved with the temporal sequ ence of trials and tested for goodness of fit with the original time series. If a brain region is actively involved in the task, the deconvolution produces an estimated response that has high and consistent amplitude and a good fit with the time series. Goodness of fit between the acquired series and the estimated time series based on deconvoltuion is tested using R2. Two types of deconvolution were performed fo r functional image analysis. First, in the response-based deconvolution, modeling of HDRs fo r each trial in which a spoken response is given commences one image prior to the image during which the response was given and is modeled for 16 images. Modeling the HDR beginni ng 1 image prior to spoken response onset allows us to see the onset of responses in area s whose activity leads up to the response; a 27 sec time course is necessary to capture delayed effects that may occur for some participants. In the second, stimulus-based deconvolut ion, modeling of the HDR commen ced at presentation of the stimulus and included 16 images. Because the in terval between stimulus onset and response can vary by as much as 4 images within a single subject, HDRs in areas whose activity is more closely associated with stimulus onset can be visualized with greater sensitivity using the stimulus-based deconvolution. In some subjects, this deconvolution adds information about frontal activity, probably because some retrieval processes can begin soon after stimulus onset. 23

PAGE 24

A single statistical param etric map was crea ted in which each voxel contained the highest of the R2 values derived from the response-based or the stimulus-based dec onvolution. To equate for sensitivity across the three different scanners used to acquire images, the top 1% of activity for each subject was used to determine if a brai n voxel showed a task-related HDR; therefore the threshold of R2 differed by subject. Similarly, to correct for brain size, the to tal number of active voxels per ROI was divided by the total amount of brain voxels for each subject. To obtain the right-hemisphere homologous ac tivity, masks were drawn over each righthemisphere ROI using the draw dataset functi on of AFNI. Subsequently, the number of active voxels under each of the masks was obtained and ente red into the analyses. Each area comprising the broad ROIs was drawn separately in order to assess their individual re lationship to degree of lesion. Five masks were drawn in the right hemis phere: pars opercularis (Pop), pars triangularis (Ptr), angular gyrus (AG), supramarginal gyrus (S MG), and the posterior third of the superior and middle temporal gyri (W+), see Table 2-2 fo r ROI names and abbreviations. The number of active voxels in Pop and Ptr was summed to derive at the total number of active voxels for the A ROI, while the number of active voxels for AG, SMG, and W+ were summed to obtain the total number of active voxels in the P ROI. For a deta iled listing of the anatom ical boundaries used to define each ROI see Appendix C. Lesion Analysis To determine degree of lesion for each left-hemisphere ROI, a modification of a lesion analysis procedure based on Naes er and colleagues was used (N aeser et al., 1998). The Naeser scale has been validated by Naes er and colleagues in several published article s and consists of the following ratings: 0 = no lesion; 1 = equivocal lesion; 2 = sma ll, patchy or partial lesion; 2.5 = patchy, less than half of area has lesion; 3 = ha lf of area has lesion; 4 = more than half of area has solid lesion; 5 = total area has solid le sion (Naeser, Palumbo, Helm-Estabrooks, Stiassny24

PAGE 25

Eder, & Martin, 1989). Two trained raters scored the brains sepa rately using the high resolution T1-weighted scans for each subject. Both raters were blinded to the subjects identity and scores on behavioral measures. Prior to rating the subjects lesions, inter-rater reliability was established using a set of four practice brains. The practice brains were from left-hemisphere stroke patients who had participated in a different study conducted in ou r laboratory. The scans we re high resolution T1 weighted images and the same rating procedures as described below we re followed. After rating each practice set independently, th e raters met to discuss any discrepancies before rating the next practice set. The correlation be tween the two sets of rati ngs for all four scans was r = .903, p < .001. To prepare the images for lesion scoring, the anterior and posterio r commissure (ac-pc) were aligned using the define ma rkers function on AFNI so they are in the same plane. Once each brain was ac-pc aligned, the raters independently scored each ROI us ing the rating scale in Appendix A. The first lesion scoring step was to anatomica lly define each of the ROIs using AFNI. The ROIs were structurally defined based on sulcal anatomy. An atlas of sulcal anatomy was referenced to accurately define each ROI and id entify inter-individual su lcal variation patterns (Ono, Kubik, & Abernathey, 1990). The templates developed by Damasio were also used for reference to help identify the ROIs (Damasio, 1995). For a list of ROI definitions see Appendix C. Each ROI was defined and rated separately by the areas that comprised them. For example, the A ROI was rated as the summed scores fo r Pop and Ptr. Similarly, the P ROI rating was 25

PAGE 26

derived from the summed ratings for AG, SMG, and W +. See Table 2-2 for a list of ROI names and abbreviations. After each rater anatomically defined the 5 RO Is, they were drawn us ing the draw dataset function of AFNI in order to identify each ROI if discrepancies between raters occurred. The slice numbers where an ROI began and ended we re noted on a scoring sh eet (see Appendix B). Subsequently, the two raters independently assessed the extent of lesion, s lice by slice, assigning a score of 0 through 5 on every slice of each RO I based on the rating scale in Appendix A. Once each ROI was scored slice by slice, a global lesion score was assigne d as the total extent of the lesion for each ROI ranging from 0 through 5 (same criteria as listed abov e). The global lesion score was based on the slice by sli ce ratings and the relative size of the lesion on each slice, but was not derived from averaging the individual slices scores. After the raters independently determined each ROIs global lesion score, the global scores were compared. Ratings that differed by one poi nt or less were averaged and used in the analyses. If a global rating differed by more than one point, the raters discussed and modified them until the 2 scores differed by one point or less, after which they were averaged. When discussing rating discrepancies, th e raters referred to the slice by slice ratings to see how they arrived at the global lesion score. Only the global lesion scores we re entered into the statistical analyses. Statistical Analyses Analysis 1 To examine the relationship between degree of lesion and right-hemisphere homologous activity, two-tailed Spearm an correlations were conducted to id entify significant relationships at the p < .05 probability level. Spear man correlations were conducted because some of the variables included in the analys is were not normally distributed. The following variables were 26

PAGE 27

correlated: degree of left-hem isphere lesion sc ores for the A and P ROIs and the number of active voxels in the right-hemisphere A and P ROIs during category member generation. Analysis 2 To examine the relationship between degree of lesion and language performance, twotailed Spearman correlations were performed betw een the A and P degree of lesion scores and the BNT and WAB-AQ components. Due to the in consistent relationshi ps reported in the literature between extent of lesi on and language performance, thes e correlations we re considered exploratory in nature. 27

PAGE 28

28 Table 2-1. Subject demographics (N=19) Mean (SD) Minimum Maximum Age 66.9 (12) 48 92 Education 13.4 (1.9) 12 18 Months since stroke 45.5 (51.9) 8 207 Notes: Ten subjects were male and 9 female; 18 were Caucasian and 1 was African American. Table 2-2. ROI names and abbreviations Anterior (A) ROI Posterior (P) ROI Pars opercularis (Pop) X Pars triangularis (Ptr) X Angular gyrus (AG) X Supramarginal gyrus (SMG) X Wernickes + (W+) X

PAGE 29

CHAP TER 3 RESULTS Due to the unavailability of fMRI data for 9 out of 19 subjects incl uded in this study, the sample sizes differed by specific aim. That is, data included in the analyses examining the association between degree of le sion and functional activity in the homologous right-hemisphere structures (Aim 1) consisted of only 10 subjects. Data includ ed in the analyses for Aim 2 included a sample size of 19 subjects. Given the small sample size and the number of comparisons performed during analyses, there was an increased risk of making a Type I error. All comparisons were corrected for family-wise e rror using Bonferroni corrections. Furthermore, since some of the variables included in the anal yses were not normally distributed, Spearmans rho correlations for rank-lev el data were conducted. To provide context for the sta tistical analyses that were c onducted in this study, Table 3-1 provides descriptive information for the fMRI activity in homologous structures of the right hemisphere during the word generation task. Tabl e 3-2 shows the mean and standard deviations of the lesion ratings for each ROI. Table 3-3 sh ows the individual lesion scores for each of the 19 subjects. Table 3-4 shows the mean and standard deviations for the language measures used to assess performance. Results for Aim 1 It was hypothesized that patient s with higher degree of lesions in the A and P regions would show more pronounced activity in right-h emisphere homologous structures due to the diminished capacity of the dominant hemisphere to provide the necessary substrates for recovery, thereby facilit ating re-lateralization of language functions to the right-hemisphere homologous structures (H eiss & Thiel, 2006). 29

PAGE 30

Two-tailed Spearm an correlations of left-hemisphere degree of lesion and homologous right-hemisphere activity reve aled no significant relations hips. Neither the A ROI ( Spearmans rho = .20, p = .57) nor the P ROI ( Spearmans rho = -.18, p = .63) degree of lesion scores were associated with their homologous functional ac tivity during the category member generation task. It is important to keep in mind that the sample size for this analysis was only 10 subjects, thus power to detect signifi cant relationships was greatly diminished. See Table 3-1 for correlation values. When examining the relationship between the homologous activity of the A and P ROIs to the BNT and WAB-AQ measures, activity in th e A ROI was negatively related to the WAB comprehension measures ( Spearmans rho = -.71, p < .05). This relationship was no longer significant after correcting for multiple compar isons. When decomposing each ROI into its respective components, the amount of ac tivity in the right-hemisphere Pop ( Spearmans rho = .69, p < .05) and AG ( Spearmans rho = -.68, p < .05) was negatively related to the WAB comprehension measures, while activity in the SMG was positively associated with the WABAQ ( Spearmans rho = .63, p = .05 ) and repetition scores ( Spearmans rho = .79, p < .01). These associations were no longer significant afte r applying Bonferroni corrections for multiple comparisons. These findings indicate that homol ogous activity in certain structures may be beneficial for performance (SMG), while activity in others may affect it (A ROI, Pop, AG). In addition, when examining the relationship between the degree of lesion for each component of the A and P ROIs to their corresponding degree of activity in the right hemisphere, no significant relationships were revealed. Ta ble 3-6 shows the values of th e correlations between homologous brain activity and behavioral la nguage measures for the sample of 10 nonfluent aphasia patients who had usable fMRI data. 30

PAGE 31

Results for Aim 2 It was hypothesized that higher de gree of lesion in the A and P ROIs would be associated with lower scores on the BNT and WAB-AQ m easures (spontaneous speech, com prehension, repetition, and naming). When using the larger sample (N=19), Spearman correlations between degree of lesion and performance on the BNT a nd WAB-AQ measures reve aled no significant associations, after Bonferroni corrections, for eith er the A or the P ROIs. Prior to correcting for multiple comparisons, the P ROI lesion score was marginally associated with WAB repetition performance (Spearmans rho = -.45, p = .06). Nevertheless, when decomposing the A and P ROIs into the indi vidual areas that comprised them, Ptr lesion score was negatively associated with WAB spontaneous speech score ( Spearmans rho = -.50, p < .05). Similarly, AG ( Spearmans rho = -.48, p < .05) and SMG lesion scores ( Spearmans rho = -.51, p < .05) were negatively associated with the WAB repetition score prior to Bonferroni corrections. These marginal asso ciations indicate that, at the individual ROI level, a greater degree of lesion is associated with poor performance. More specifically, anterior regions (P tr) are associated with nonfluen cy, while a greater degree of lesion in posterior perisylvian ar eas (AG & SMG) is associated w ith poorer repetition ability. For a listing of correlations between degree of lesi on and the BNT and WAB-AQ measures see Table 3-7. Refer to Tables 3-2 through 3-5 for descriptive information regarding lesion scores, activity levels, and behavioral performance. 31

PAGE 32

Table 3-1. C orrelations between Anterior and Posterior ROI lesion scores and homologous activity (N = 10) Anterior score Anterior activity Posterior score Posterior activity Anterior score 1 0.2 0 -0.07 Anterior activity 0.2 1 -0.09 0.07 Posterior score 0 -0.09 1 -0.18 Posterior activity -0.07 0.07 -0.18 1 Notes: Two-tailed Spearmans correlations. N one were significant. Functional activity was derived from the right-hemisphere homologous ROIs during category member generation task. Anterior ROI = sum of pars triangularis and pa rs opercularis. Posterior ROI = sum of angular gyrus, supramarginal gyrus, and wernickes +. Table 3-2. Number of right-hemis phere active voxels (N = 10) Mean SD Minimum Maximum Anterior ROI 399.3 949.2 29 3090 Posterior ROI 453.4 349.71 64 1261 Pars opercularis 223.51 550.65 0 1788.12 Pars triangularis 175.76 400.41 0 1301.52 Angular gyrus 35 33.65 0 89.16 Supramarginal gyrus 31.06 30.27 0 83.9 Wernickes + 387.44 348.84 43.48 1227.95 Notes: Functional activity was derived from th e right-hemisphere homologous structures during category member generation task. Anterior ROI = su m of pars triangularis and pars opercularis. Posterior ROI = sum of angular gyrus, s upramarginal gyrus, and wernickes +. Table 3-3. Left-hemisphere lesion ratings (N= 19) Mean SD Minimum Maximum Anterior ROI 4.55 3.55 0 10 Posterior ROI 5.21 3.77 0 11.5 Pars opercularis (Pop) 2.83 1.74 0 5 Pars triangularis (Ptr) 1.72 2.08 0 5 Angular gyrus (AG) 1.32 1.33 0 4 Supramarginal gyrus (SMG) 2.26 1.55 0 4.5 Wernickes + (W+) 1.63 1.52 0 4.5 Notes: Descriptives for each left-hemisphere ROI. Anterior ROI = sum of Pop and Ptr. Posterior ROI = sum of AG, SMG, and W+. Maximum lesion rating for Anterior ROI = 10 (5x2). Maximum lesion rating for Posterior ROI = 15 (5x3). 32

PAGE 33

33 Table 3-4. Individua l lesion ratings Subject Anterior ROI Posterior ROI Pop Ptr AG SMG W+ 1 9 2.5 4.5 4.5 0.5 2 0 2 10 6.5 5 5 1 4.5 1 4 5.5 1 4 1.5 0 1 0 5 4 4.5 4 0 1 2.5 1 6 5.5 4.5 4 1.5 0 2 2.5 7 0 0.5 0 0 0 0 0.5 8 1.5 11.5 1.5 0 4 3.5 4 9 0 8.25 0 0 2.75 4 1.5 10 2.5 0 2.5 0 0 0 0 11 5.5 1 2.75 2.75 0 0 1 12 3.5 9 3 0.5 2 3.5 3.5 13 10 8.25 5 5 2.25 4.5 1.5 14 0.5 11.5 0.5 0 3.5 3.5 4.5 15 7.5 10 3 4.5 2.5 3.5 4 16 2 7.5 2 0 2.5 2.5 2.5 17 3 2.5 3 0 0 2.5 0 18 10 3.5 5 5 1 2.5 0 19 0 4.5 0 0 2 0 2.5 21 6.5 2 4 2.5 0 1 1 Notes: Degree of lesion ratings for each of the 19 subjects included in the study. Individual ROI scores are the averaged global scores of each rate r. Anterior degree of lesion was calculated as the summed scores of Pop and Ptr (Brocas Area). The posterior degree of lesion was calculated as the summed scores of AG, SMG and W+. Pop = pars opercularis; Ptr = pars triangularis; AG = angular gyrus; SMG = supramargina l gyrus; W+ = wernickes plus. Table 3-5. Behavioral language measures (N = 19) Mean SD Minimum Maximum BNT % correct 0.44 0.21 0.13 0.83 WAB Aphasia Quotient 67.31 11.38 45.4 81.9 WAB spontaneous speech 12.26 3.02 7 16 WAB comprehension 170.26 15.50 135 200 WAB repetition 61.42 22.85 24 98 WAB naming 67.32 16.17 31 92 Notes: Scores on the BNT were expressed as pe rcent correct (out of 60 items) while scores on the WAB were expressed as the raw scores for each component of the AQ. Higher scores indicate better performance. BNT = Boston Naming Test; WAB = Western Aphasia Battery.

PAGE 34

Table 3-6. Correlations between ROIs, homologous activity, and language performance (N = 10) 34 1 2 3 4 5 6 7 8 9 10 1. Pop score 1 2. Pop activity 0.20 1 3. Ptr score 0.80** 0.22 1 4. Ptr activity -0.01 0.69* 0.02 1 5. AG score -0.05 -0.24 -0.18 -0.56 1 6. AG activity 0.25 0.47 0.54 0.63 -0.56 1 7. SMG score 0.26 -0.18 0.12 -0.44 0.86** -0.23 1 8. SMG activity -0.02 -0.21 -0.22 -0.21 -0.33 -0.49 -0.53 1 9. W+ score -0.25 0.34 -0.20 0.07 0.54 -0.06 0.56 -0.54 1 10. W+ activity -0.20 0.08 -0.01 -0.12 -0.10 -0.12 -0.32 0.51 -0.01 1 11. BNT % correct 0.32 0.18 0.01 0.13 -0.24 -0.15 -0.05 0.41 0.03 -0.18 12. WAB Aphasia Quotient 0.10 0.02 -0.12 -0.08 -0.27 -0.37 -0.26 0.63* -0.13 -0.04 13. WAB spontaneous speech 0.17 0.41 -0.09 0.31 -0.38 -0.08 -0.29 0.31 0.01 -0.31 14. WAB comprehension -0.37 -0.69* -0.52 -0.47 0.08 -0.68* -0.06 0.60 -0.13 0.12 15. WAB repetition -0.13 0.05 -0.37 0.16 -0.51 -0.23 -0.54 0.79** -0.26 0.24 16. WAB naming 0.18 0.10 0.40 -0.16 -0.35 0.05 -0.31 0.16 -0.17 -0.10 Notes: Two-tailed Spearm ans correlations between individual ROIs, right-hemisphere ho mologous activity, and behavioral languag e measures. Correlation is signifi cant at the .05 level. ** Correl ation is significant at the .01 level. All correlations were no longer significant following family-wise B onferroni corrections (required p < .01 for significance on correlati ons between ROI lesion scores and homologous activity; required p < .001 for significance on correlations between le sion scores, homologous activity and language performance measures). Pop = pars opercularis; Ptr = pars tr iangularis; AG = angular gyrus; SM G = supramarginal gyrus; W+ = wernickes plus; BNT = Boston Naming Test; WAB = Western Aphasia Battery.

PAGE 35

Table 3-7. Correlations between lesion scores and language performance (N = 19) 1 2 3 4 5 6 7 1. Anterior score 1 2. Posterior score -0.08 1 3. Pop score 0.94** -0.11 1 4. Ptr score 0.93** -0.03 0.81** 1 5. AG score -0.23 0.91** -0.28 -0.15 1 6. SMG score 0.23 0.82** 0.25 0.23 0.71** 1 7. W+ score -0.30 0.84** -0.39 -0.19 0.74** 0.45 1 8. BNT % correct 0.09 -0.03 0.10 -0.02 -0.12 -0.01 -0.01 9. WAB Aphasia Quotient -0.26 -0.37 -0.20 -0.42 -0.41 -0.45 -0.20 10. WAB spontaneous speech -0.32 -0.18 -0.21 -0.50* -0.25 -0.41 0.04 11. WAB comprehension -0.01 -0.07 -0.05 -0.11 0.04 0.11 -0.14 12. WAB repetition -0.13 -0.45* -0.15 -0.24 -0.48* -0.51* -0.23 13. WAB naming -0.11 -0.15 -0.08 -0.14 -0.23 -0.18 -0.12 35 Notes: Two-tailed spearman correlations between degree of le sion scores and behavi oral language measur es. Correlation is significant at the .05 level. ** Correlation is significant at the .01 leve l. All correlations were no longer significant follo wing familywise Bonferroni corrections (required p < .001 for significance). Pop = pars opercularis; Ptr = pars triangularis; AG = angular gyrus; SMG = supramarginal gyrus; W+ = wernickes plus; BNT = Boston Naming Test; WAB = Western Aphasia Battery.

PAGE 36

CHAP TER 4 DISCUSSION This chapter reviews the major findings of th e current study and discusses their theoretical implications. It also focuses on discussing the lim itations of the current st udy and suggests future directions to expand the field of lesion analysis and recovery from aphasia. It is important to restate that the current study had a relatively sma ll sample size (N=19), and therefore the results should be interpreted with caution. Review of Findings Review of Aim 1 It was expected that patients with higher de gree of lesions in the A and P regions would show more pronounced activity in right-hemisphere homologous structures due to the diminished capacity of the dominant hemisphere to provide the necessary substrates for recovery, thereby facilitating re-lateralization of language functions to the right-hemisphere homologous structures (Heiss & Thiel, 2006). Results i ndicated that there was no associ ation between degree of lesion and the amount of right-hemisphere homologous ac tivity in the A and P ROIs. Therefore, the current study was unable to replic ate and extend the work of Blank and colleagues (Blank et al., 2003), suggesting that right-hemisphere structur es (pars opercularis) show more pronounced activity only when their left-hemisphere coun terparts are damaged. There are two possible explanations for why the current study could not replicate Blanks findings: (1) sample characteristics and (2) imaging task. Blanks sa mple consisted of 14 previous nonfluent aphasia patients who had recovered the ability to speak spontaneously in sentences, while the current study included subjects who were currently nonf luent. Moreover, Blank used PET to image language functions by having subjec ts respond to an autobiographica l question such as tell me what you like to do on a holiday, while the curre nt study implemented a category member word 36

PAGE 37

generation task, which m ay activate different areas of the language network. The fact that the samples and imaging tasks were so different coul d have affected the results of the current study. Similarly, it is possible that the use of three different 3T scanners to collect data added significant error variance due to differential sensitivity to blood-oxygen-level dependent (BOLD) changes. On the other hand, when analyses of each of the 5 regions that comprised the A and P ROIs were examined, the amount of functiona l activity in the right-hemisphere SMG was positively associated with the WAB-AQ and repe tition scores, indicating that more pronounced right-hemisphere activity in the SMG was relate d to less severe aphasi a and better repetition abilities. These findings are inline with previous studies suggesting that ri ght-hemisphere activity may help to compensate for the loss of langua ge functions during rehabilitation in aphasia (Crinion & Price, 2005). In contra st, activity in right Pop and AG was negatively related to WAB comprehension scores indicating that more pronounced activity in these right-hemisphere structures was detrimental to performance on comprehension abilities. These results are similar to previous findings suggesting th at heightened right-hemisphere ac tivity in some structures can hinder recovery in nonfluent aphasia patients rath er than enhancing it (Martin et al., 2004). The overall findings point to the importance that damage to certain left-hemisphere structures plays in language performance for nonflue nt aphasia. They also suggest that there is a beneficial role that certain right-hemisphere structures (such as the SMG) may play in the performance of different aspects of language functioning. Although the main hypothesis was not supported, the marginal relationships mentioned above revealed that functional activity in specific regions of the right hemisphere wa s related to both good and poor performance on language tasks. 37

PAGE 38

In sum these findings revealed the complexity of the relationship between lesion patterns, right-hemisphere activity, and la nguage performance in aphasia that have been previously observed in the literatu re (Crinion & Leff, 2007; Crosson et al., 2007) and suggest that, while posterior right-hemisphere activity may enhance performance on repetition and mitigate overall aphasia severity, activity in nei ghboring structures might be detrim ental to recovery of auditory comprehension. It can be concluded that Blanks findings are not universal and cannot be applied to this sample. Review of Aim 2 It was expected that higher degree of lesion in the A and P ROIs would be associated with lower scores on the BNT and WAB-AQ measures (spontaneous speech, comprehension, repetition, and naming). Results revealed margina lly significant relations hips between anterior areas and overall fluency scores and posterior regions and repe tition abilities, all of which indicated that more severe lesions were rela ted to poorer performance. More damaged Ptr was related to worse performance on spontaneous sp eech, consistent with previous literature suggesting that speech production is related to more left frontal activity, which indeed necessitates more intact structures to function properly (Crinion & Leff, 2007). Angular gyrus and SMG, both structures of the inferior parietal lobul e, have previously been linked to language performance in aphasi a (Kertesz et al., 1993), which was supported by this study since higher degree of lesion in these two areas was associated with poorer repetition abilities. Although many studies have reported relations hips between lesion characteristics and specific language deficits, others have argued that lesion -deficit analysis is an uncertain method of defining the location of the neural sub-systems involved in language processing (Wise, 2003) due to the variability of cortical organization of the language netw ork at the individual level, and 38

PAGE 39

the fact that dam age resulting from stroke is extremely different from person to person, where damage to cortex, white matter, and subcortical structures vary greatly. Subjects in this study had very different lesions. Some lesions were large and solid; others were small and patchy, while some subjects had a greater degree of lesion in the an terior than posterior re gions or vice versa. It is possible that different result s could have been obtained if th e current study included subjects with very similar lesion characteristics, in which case inferences regarding specific deficits could have been more transparent. In sum, only marginal relationships were found between degree of lesion and performance on different language measures. This may be attributed to the vari ability of lesion distributions included in this sample which can affect the co nnections between different areas of the language network for some subjects and not others, thus obscuring the relationshi p between location and degree of lesion and specific deficit. Study Implications As stated earlier, the literature is mixed with regards to the relationship between left and right-hemisphere activity and la nguage performance in aphasia. Some studies have been conducted that investigated this relationship; however, they ha ve included different types of aphasia syndromes which makes it difficult to generalize the findings to nonfluent aphasia patients. When scientists bette r understand the reorganization of language functions in aphasia and how it relates to lesion characteristics and re covery of language functi ons, they will be better equipped to develop rehabilitation tr eatments that target the re-activation of either left or righthemisphere structures while decreasing activity considered detrimental for recovery. This study was unable to determine the relationship between degree of lesion and right-hemisphere activity in the homologous structures of damaged left-hem isphere areas; however a larger sample with more circumscribed lesion distri butions could shed much needed light into this relationship. 39

PAGE 40

Sim ilarly, it is crucial to bette r understand the lesion correlates of specific language deficits to advance our understanding of the neural su bstrates of language production in nonfluent aphasia. Better understanding of th is relationship could aid therapis ts in predicting what kinds of deficits to expect from specific lesion patte rns, and how these patterns can predict both spontaneous and treatment-induced recovery. If therapists can predict the lesion patterns associated with recovery from specific therapies, they will be better able to maximize recovery and minimize health care costs by selecting treatm ents with a higher prob ability of success in patients with specific lesion patter ns and associated deficits. Study Limitations There were several limitations to this study. First and foremost the sample size for Aim 1 consisted of only 10 subjects who had available fMRI data at the time the analyses were conducted. Small samples decrease statistical power to detect significant relationships above chance levels. A second limitation of this study was that the subjects data were collected from two slightly different studies. There were 8 subjects from study 1 and 11 from study 2. Another aspect that differed between studies 1 and 2 was the presenta tion of the stimuli during the fMRI task, which could have affect ed functional activity patterns in the right hemisphere. Both studies used the same imagi ng task (category member generation); however study 1 presented the stimuli audito rily, while stimuli presentation for study 2 was audio-visual. Moreover, structural and functiona l data were acquired using three different 3T scanners, which could have affected the sensitivity to detect task-related brain changes by study. To equate for sensitivity across the three scanne rs, only the top 1% of task-rel ated functional activity for each subject was analyzed. A third limitation of this study was that it did not examine the conn ectivity between areas of the language network, which is important to the reactivation of st ructures involved in 40

PAGE 41

41 recovery, especially since damage from stroke is so variable and may include white matter lesions in certain subjects and not others, thus causing different behavioral deficits (Wise, 2003). Future Directions The study of recovery in apha sia is complex and remains poorly understood. To further investigate the relationship between degree of lesion, right-hemisphere activity, and related deficits, longitudinal studies in cluding only one aphasia syndrome, similar behavioral deficits, and circumscribed lesion characteristics shou ld be undertaken. Conducting such a study would likely be arduous due to the inter-individual va riability of lesion characteristic and related deficits observed after stroke. Ideally, treatmen t studies will be carried out which take into consideration lesion characteristics, functional activ ity, and behavioral defi cits before and after language treatment, as well as at long-term follow-up to measur e the longevity of effects. Furthermore, future studies should examin e the integrity of white matter pathways connecting areas of the language network to one another to further under stand the reorganization of language functions after stroke. Techniques such as diffusion tensor imaging (DTI) can be implemented in conjunction with fMRI to examin e the connectivity between functional areas of the language network as well as assessing change s in white matter integrity over the course of treatment and their relation to functional reorganization (Bihan et al., 2001). Conducting such a study may prove a monumental task, but it would provide much needed information regarding recovery in aphasia and poten tially aid in the development and implementation of novel treatments aiming to enhance recovery in pa tients with specific deficits and lesion characteristics.

PAGE 42

APPENDIX A DEGR EE OF LESION RATING SCALE 42

PAGE 43

APPENDIX B DEGREE OF LESION SCORING SHEET 43

PAGE 44

APPENDIX C ROI DEFINITIONS 1) Pars Opercularis (Pop)/ BA 44: Anterior border = anterior ascending ramus of the sylvian fissure Posterior border = inferior precentral sulcus (anterior subcentral sulcus) Superior border = inferior frontal sulcus Inferior border = main br anch of the sylvian fissure In the depth of the sylvian fissure it borders on the insula. 2) Pars Triangularis (Ptr)/ BA 45: Anterior/superior border = inferior frontal sulcus Posterior border = anterior ascen ding ramus of the Sylvian fissure Inferior border = anterior horiz ontal ramus and the sylvian fissure Borders on the insula in the depth of the sylvian fissure. 3) Angular Gyrus (AG)/ BA 39: In parietal lobe Surrounds the posterior end of the superior temporal sulcus follow posterior portion of the superior temporal gyrus into pa rietal lobe (go medial to define the depth of the posterior ascending ramu s of the sylvian fissure it has to connect) Bounded anteriorly by the supramarginal gyrus Bounded posteriorly by the parieto-occipital sulcus 4) Supramarginal Gyrus (SMG)/ BA 40: Found on either side of the ascending ramus of the sylvian fissure Anterior border = inferior portion of the postcentral sulcus Posterior border = angular gyrus Superior border = intraparietal sulcus 5) Wernickes + (W+): Posterior third of the superior temporal gyrus and middle temporal gyrus Superior Temporal Gyrus (BA 22): Anterior border = find the most anterior point of Heschl gyrus and draw a vertical line down to the superior temporal sulcus Posterior border = draw a vertical line from the upper inflection of the posterior ascending ramus of the sylvian fissure until the end of the superior temporal suclus Superior border: sylvian fissure Inferior border = superior temporal sulcus Middle Temporal Gyrus (BA 21): Anterior border = find the most anterior point of Heschl gyrus and draw a vertical line down to the inferior temporal sulcus Posterior border = draw a vertical line from the upper inflection of the superior temporal suclus until the end of the inferior temporal suclus Superior border = superior temporal sulcus Inferior border = inferior temporal sulcus 44

PAGE 45

LIST OF REFERE NCES Abo, M., Senoo, A., Watanabe, S., Miyano, S., Dose ki, K., Sasaki, N., et al. (2004). Languagerelated brain function during word re petition in post-stroke aphasics. Neuroreport, 15 (12), 1891-1894. Barlow, T. (1877). On a case of double cerebral he miplegia, with cerebral symmetrical lesions. British Medical Journal, 2 103-104. Basso, A., & Farabola, M. (1997) Comparison of Improvement of Aphasia in Three Patients with Lesions in Anterior, Posterior, and Antero-posterior Language Areas. Neuropsychological Rehabilitation, 7 (3), 215-230. Basso, A., Gardelli, M., Grassi, M. P., & Mariotti, M. (1989). The role of the right hemisphere in recovery from aphasia: Two case studies. Cortex, 25, 555-566. Bihan, D. L., Mangin, J. F., Poupon, Clark, C. A., Pappata, S., Molko, N., et al. (2001). Diffusion Tensor Imaging: Concepts and Applications. Journal of Magentic Resonance Imaging, 13 534. Blank, S. C., Bird, H., Turkheimer, F., & Wise, R. J. (2003). Speech production after stroke: the role of the right pars opercularis. Annals of Neurology, 54 (3), 310-320. Borovsky, A., Saygin, A. P., Bates, E., & Dronkers, N. (2007). Lesion correlates of conversational speech production deficits. Neuropsychologia, 45 (11), 2525-2533. Breier, J. I., Castillo, E. M., Boake, C., Billingsley, R., Maher, L., Francisco, G., et al. (2004). Spatiotemporal patterns of language-specific br ain activity in patients with chronic aphasia after stroke using magnetoencephalography. Neuroimage, 23(4), 1308-1316. Cao, Y., Vikingstad, E. M., George, K. P., Johns on, A. F., & Welch, K. M. (1999). Cortical language activation in stroke patients recovering from a phasia with functional MRI. Stroke, 30(11), 2331-2340. Cornelissen, K., Laine, M., Tarkiainen, A., Ja rvensivu, T., Martin, N., & Salmelin, R. (2003). Adult brain plasticity elicited by anomia treatment. Journal of Cognitive Neuroscience, 15(3), 444-461. Cox, R. W. (1996). AFNI: Software for analysis and visualization of functional magnetic resonance images. Computers in Biomedical Research, 29, 162-173. Crinion, J., & Price, C. J. (2005). Right anterior superior temporal activation predic ts auditory sentence comprehension following aphasic stroke. Brain, 128(Pt 12), 2858-2871. Crinion, J. T., & Leff, A. P. (2007). Recovery an d treatment of aphasia af ter stroke: functional imaging studies. Current Opinion in Neurology, 20 667. 45

PAGE 46

Crosson, B., McGregor, K., Gopinath, K., Conw ay, T., Benjam in, M., Chang, Y.-L., et al. (2007). Functional MRI of Language in Aphasi a: A Review of the Literature and the Methodological Challenges. Neuropsychology Review, 17(2), 157-177. Crosson, B., Moore, A. B., Gopinath, K., White, K. D., Wierenga, C. E., Gaiefsky, M. E., et al. (2005). Role of the right and le ft hemispheres in recovery of function during treatment of intention in aphasia. Journal of Cognitive Neuroscience, 17 (3), 392-406. Damasio, H. (1995). Human brain anatomy in computerized images. New York: Oxford University Press. Dronkers, N. F., Wilkins, D. P., Van Valin, R. D ., Jr., Redfern, B. B., & Jaeger, J. J. (2004). Lesion analysis of the brain areas involved in language comprehension. Cognition, 92 (12), 145-177. Duffau, H., Bauchet, L., Lehericy, S., & Capelle, L. (2001). Functional comp ensation of the left dominant insula for language. Neuroreport, 12 (10), 2159-2163. Dunn, L. M., & Dunn, D. M. (2007). The Peabody Picture Vocabulary Test, Fourth Edition Bloomington, MN: NCS Pearson, Inc. Gold, B. T., & Kertesz, A. (2000). Right hemisphere semantic processing of visual words in an aphasic patient: an fMRI study. Brain and Language, 73 (3), 456-465. Goodglass, H., & Kaplan, E. (1983). The Assessment of Aphasia and Related Disorders (Second ed.). Philadelphia: Lea & Febiger. Gopinath, K., Crosson, B., McGregor, K., Peck, K. K., ChangY.-L., Moore, A., Sherod, M., Cavanagh, C., Wabnitz, A., Wierenga, C., White, K., Cheshkova, S., Krishnamurthy, V., Briggs, R. W. (in press). Selective detren ding method for reducing task-correlated motion artifact during speech in event-related FMRI. Human Brain Mapping. Heiss, W. D., Karbe, H., Weber-Luxenburger, G., Herholz, K., Kessler, J., Pietrzyk, U., et al. (1997). Speech-induced cerebral metabolic activation reflects recovery from aphasia. J Neurol Sci 145(2), 213-217. Heiss, W. D., Kessler, J., Thiel, A., Ghaemi, M., & Karbe, H. (1999). Differential capacity of left and right hemispheric areas for co mpensation of poststroke aphasia. Annals of Neurology, 45(4), 430-438. Heiss, W. D., & Thiel, A. (2006). A proposed re gional hierarchy in recovery of post-stroke aphasia. Brain and Language, 98 (1), 118-123. Kaplan, E. F., Goodglass, H., & Weintraub, S. (1983 ). The Boston Naming Test Philadelphia, PA: Lea & Febiger. 46

PAGE 47

Karbe, H., Thiel, A., W eber-Luxenburger, G., Her holz, K., Kessler, J., & Heiss, W. D. (1998). Brain plasticity in poststroke aphasia: what is the contribution of the right hemisphere? Brain and Language, 64(2), 215-230. Kertesz, A. (1982). Western Aphasia Battery Orlando, FL: Grune & Stratton. Kertesz, A., Lau, W. K., & Polk, M. (1993). The Structural Determinants of Recovery in Wernicke's Aphasia. Brain and Language, 44 (2), 153-164. Kertesz, A., & McCabe, P. (1977). Recovery Patterns and Prognosis in Aphasia. Brain, 100 (1), 1-18. Kinsborne, M. (1971). The minor cerebral hemisphere as a source of aphasic speech. Archives of Neurology, 25 302-306. Knopman, D. S., Selnes, O. A., Niccum, N., & R ubens, A. B. (1984). Recovery of naming in aphasia: Relationship to fluency, comprehension and CT findings. Neurology, 34 14611470. Kreisler, A., Godefroy, O., Delmaire, C., Debac hy, B., Leclercq, M., Pruvo, J.-P., et al. (2000). The anatomy of aphasia revisited. Neurology, 54 (5), 1117-1123. Lazar, R. M., Marshall, R. S., Pile-Spellman, J., Duong, H. C., Mohr, J. P., Young, W. L., et al. (2000). Interhemispheric transfer of language in patients with left frontal cerebral arteriovenous malformation. Neuropsychologia, 38 1325-1332. Leger, A., Demonet, J. F., Ruff, S., Aithamon, B. Touyeras, B., Puel, M., et al. (2002). Neural substrates of spoken language rehabilita tion in an aphasic patient: an fMRI study. Neuroimage, 17(1), 174-183. Martin, P. I., Naeser, M. A., Theoret, H., Tormos J. M., Nicholas, M., Ku rland, J., et al. (2004). Transcranial magnetic stimulation as a complementary treatment for aphasia. Seminars on Speech and Language, 25 (2), 181-191. Miura, K., Nakamura, Y., Miura, F., Yamada, I., Takahashi, M., Yoshikawa, A., et al. (1999). Functional magnetic resonance imaging to word generation task in a patient with Broca's aphasia. Journal of Neurology, 246 (10), 939-942. Naeser, M., Palumbo, C. L., Helm-Estabrooks, N., Stiassny-Eder, D., & Martin, A. (1989). Severe nonfluency in aphasia: Role of the medial subcallosal fasciculus and other white matter pathways in recovery of spontaneous speech. Brain, 112(1), 1-38. Naeser, M., Palumbo, C. L., & Stiassny-Eder, D. (1990). Late recovery of auditory comprehension in global aphasia:Improved reco very observed with subcortical temporal isthmus lesion vs. Wernicke 's cortical area lesion. Archives of Neurology, 47 425-432. 47

PAGE 48

48 Naeser, M. A., Baker, E. H., Palumbo, C. L., Nic holas, M., Alexander, M. P., Samaraweera, R., et al. (1998). Lesion Site Patterns in Severe, Nonverbal A phasia to Predict Outcome With a Computer-Assisted Treatment Program. Archives of Neurology, 55 (11), 1438-1448. Naeser, M., Theoret, H., Kobayashi, M., Martin, P., Nicholas, M., et al. (2002). Modulation of cortical areas with repetitive transcranial m agnetic stimulation to improve naming in nonfluent aphasia [Abstract #133]. 8th Inte rnational Conference on F unctional Mapping of the Human Brain, June 2-6, 2002, Sendai, Japan. Available on CD-ROM in Neuroimage 16(2). Ono, M., Kubik, S., & Abernathey, C. D. (1990). Atlas of the cerebral sulci New York: Thieme Medical Publishers, Inc. Seghier, M., Lazeyras, F., Momjian, S., Annoni, J. M., de Tribolet, N., & Khateb, A. (2001). Language representation in a patient with a dominant right hemisphere: fMRI evidence for an intrahemispheric reorganisation. Neuroreport, 12(13), 2785-2790. Selnes, O. A., Knopman, D. S., Niccum, N., Rubens, A. B., & Larson, D. (1983). Computed tomographic scan correlates of auditory comp rehension deficits in aphasia: A prospective recovery study. Annals of Neurology, 13 (5), 558-566. Warburton, E., Price, C. J., Swi nburn, K., & Wise, R. J. (1999). Mechanisms of recovery from aphasia: evidence from positron emission tomography studies. Journal of Neurology, Neurosurgery, and Psychiatry, 66 (2), 155-161. Weiller, C., Isensee, C., Rijntjes, M., Huber, W., Muller, S. P., Bi er, D., et al. (1995). Recovery from Wernicke's Aphasia: A pos itron emission tomographic study. Annals of Neurology, 37, 723-732. Wise, R. J. S. (2003). Language systems in nor mal and aphasic human subjects: functional imaging studies and inferences from animal studies. Br Med Bull, 65(1), 95-119.

PAGE 49

BIOGR APHICAL SKETCH Zvinka Zlatar received her Bachelor of Arts degree in psychology from the University of Nevada, Las Vegas in 2005 and her Master of Scie nce degree in psychology from the University of Florida in the spring of 2009. As a post-baccal aureate, she worked at the University of Florida's NIMH Center for the Study of Em otion and Attention and the Brain Imaging Rehabilitation & Cognition Lab. Zvinka is curr ently a second-year neuropsychology graduate student in the University of Florida's Clinical and Health Psychology program. Her research focuses primarily on studying cognition in old ag e and how the brain-behavior relationship becomes affected after stroke and degenera tive disease, as well as in healthy aging. Recently, Zvinka has been involved in an in tention-based language treatment study aiming to improve language production in individuals wi th nonfluent aphasia. Currently, she is involved in conducting an fMRI study that will investigat e brain activity and cognitive differences as a function of fitness level in sedentar y and fit younger and older adults. 49