Neural signatures of verb argument structure in agrammatic aphasic and age-matched individuals B. Bonakdarpour a, * , C.K. Thompson a,b , S.C. Fix a a Aphasia and Neurolinguistics Laboratory, Department of Communication Sciences and Disorders, Northwestern University, 2240 Campus Drive, Evanston, IL 60208, USA b Cognitive Neurology and Alzheimer’s Disease Center, Northwestern University, USA Background Cross-linguistic research on verb processing in agrammatic aphasia has shown that verb production is influenced by verb argument structure com- plexity, i.e., verbs with more complex argument structure configurations are more difficult to produce than those with simpler lexical entries (e.g., De Bleser & Kauschke, 2003; Kim & Thompson, 2004; Jonkers & Bastia- anse, 1998; Luzzatti et al., 2002). In a recent neuroimaging study Thomp- son et al., in press also found an argument structure effect in healthy normals, with increased activation of the posterior perisylvian network associated with verbs with more complex argument structure . The pur- pose of this study was to investigate the neural networks involved in verb argument structure processing in a group of agrammatic aphasic patients and an age-matched controls. Methods Participants Five agrammatic aphasic individuals (4 male and 1 female) and 14 age- matched controls (mean age, 56 and 54 years, respectively) participated in the study. Agrammatic aphasia was diagnosed based on the Western Aphasia Battery (WAB, Kertesz, 1982) as well as spontaneous speech analysis and other tests. The etiology of aphasia was stroke in all patients. FMRI stimuli and procedures Stimuli included 40 one-, two-, and three-argument verbs as well as animal and tool names pseudowords, which were visually presented for a lexical decision using an event-related fMRI design. Unaccusative, psych, complement and morphologically zero-derived verbs were excluded. Stimuli were pseudorandomized using OPTSEQ and were pre- sented in two runs of 6 min and 4 s each using SuperLab (Cedrus Corp., version 2.00 for PC, Phoenix, Arizona). Each word or pseudoword was displayed for 1200 ms followed by a 500 ms inter-trial interval. Null events of 1700 or 3400 ms were randomly interspersed in each run. Functional MRI scans A 3T Siemens scanner was used for anatomical and functional scans. Functional volumes were acquired with the following parameters: TR 2 s, TE 30 ms flip angle 90°, FOV 220 mm, matrix 64 · 64 and slice thick- ness 3 mm. An MP-RAGE sequence with following parameters was used for anatomical scans: TR/TE of 2100 ms/2.4 ms, flip angle of 8°, TI of 1100 ms, matrix size of 256 mm · 256 mm; FOV of 22 cm, and slice thick- ness of 1 mm. Data analysis SPM2 (Welcome Department of Imaging Neuroscience, University College London) was used for data analysis. Functional scans were cor- rected for slice-acquisition timing and realigned to a mean functional image. The anatomical volume was co-registered to the mean image, and normalized to the MNI 152-subject template brain. For patients, SPM’s ImCalc function was used to create a binary mask of the anatom- ical image with lesioned cortex = 0 and healthy cortex = 1. The functional volumes were then normalized using the same transformation and were smoothed using a 10 mm FWHM isotropic Gaussian kernel. Results Control participants Main effects for verbs and nouns, for age-match controls, showed bilateral activation of neural networks known to be involved in word pro- cessing. Analysis of contrasts related to verb type showed an argument structure effect, similar to that observed in young normals. The most pro- nounced difference was observed when 3-place verbs were compared to 1- place verbs. This contrast showed a significant cluster located in the left angular gyrus and was significant at p = 0.001, corrected for multiple com- parisons (Fig. 1, top row). Aphasic individuals Aphasic patients, like age-matched participants showed bilateral recruitment of language networks for verbs and nouns, except of course doi:10.1016/j.bandl.2007.07.056 * Corresponding author. E-mail address: borna@northwestern.edu (B. Bonakdarpour). www.elsevier.com/locate/b&l Brain and Language 103 (2007) 8–249