Objective The aim of this study was to evaluate brain connectivity by diffusion tensor imaging (DTI) in schizophrenia patients with clozapine‐induced obsessive compulsive symptoms (OCS). Methods Eighteen schizophrenia patients, nine of which had clozapine‐induced OCS (Clz‐OCS (+)), 9 without OCS (Clz‐OCS (−)) and 9 healthy controls were included. Psychopathology was evaluated with Positive and Negative Syndrome Scale and Yale‐Brown Obsession and Compulsion Scale in the patient groups. All groups were assesed with neurocognitive tests and DTI. Results Tract‐Based Spatial Statistics based comparison of DTI revealed lower fractional anisotropy in the genu of corpus callosum (CC), right cingulum, left frontal white matter (WM) in the Clz‐OCS (+) group, compared to controls. Fractional anisotropy was found to be lower in the bilateral occipital WM and higher in the bilateral medial temporal regions, anterior limb of internal capsule, cingulum, frontoparietal peripheral WM, right external capsule and genu of CC in Clz‐OCS (+) patients compared to Clz‐OCS (−). Conclusions WM integrity in several pathways such as cortico‐striato‐thalamo‐cortical circuitry and orbito‐frontal tracts seems to be affected differently in patients with Clz‐OCS (+). Different neuroplastic effects of clozapine leading to occurrence of OCS in a subgroup of patients is possible, and needs further evaluation by longitudinal follow‐up studies.

Background: The Broadman Area 17 (V1) has a good representation of retinotopic map. Similarity between visual input and the representation of it in V1 would be affected from both an intrinsic noise and the saccadic eye movements. GABA’s role in increasing signal to noise ratio is known but, how GABAergic activity helps to control noise, based on input and saccades, has not been modelled. Methods: A computational model of V1 was designed by using the MATLAB 2021a platform, and different six images, each containing a circle, triangle, and square, were used to test the model. The developed V1 was constituted of six different orientation columns (OCs). Each OC contains GABAergic and glutamatergic connections. Thus, OCs were activated not only based on afferent image inputs but also on the interaction among fired columns via the sum of glutamate and GABAergic neuron weights. V1 representation states for twelve, twenty and thirty saccades were summed and visualized. Finally, the original and representational forms of the image were compared. In the model, GABA activity levels have been tuned and the results of each level analysed. Results: It has been shown that level of GABA activity in the orientation columns during saccades is a critical factor for an ideal image representation. Decreased levels of GABA activity can be associated with inadequacy of noise elimination which could impair correct contour perception. Conclusion: Orientation columns can be conceptualized as microprocessors of V1. In this region, images are represented with high similarity. This similarity seems to need efficient GABAergic activity.