Spatial heterogeneity of functional Magnetic Resonance Imaging indices of dorsolateral prefrontal cortex activation evoked by a working memory task: A comparison of patients with schizophrenia and healthy controls

Background: This study is based on the hypothesis that functional Magnetic Resonance Imaging (fMRI) indices in the dorsolateral prefrontal cortex (DLPFC) of schizophrenia patients are spatially more heterogeneous than healthy controls for activation evoked by a working memory (WM) task. Patients have shown to have greater activation than controls in the DLPFC, but this seems to cancel out when performing group averages, which could be explained by patients having more spatially distributed activation. This may one of the causes for discrepant findings concerning hypo- or hyperactivation in the DLPFC of patients when performing a WM task. Methods: The cohort consisted of demographically matched schizophrenia patients and healthy controls. fMRI data was acquired to study the activation evoked by a modified Sternberg Item Recognition Paradigm (SIRP) known to induce robust activation of the main brain areas subserving WM both in schizophrenia patients and healthy controls. Those areas are the DLPFC, the intraparietal sulcus, the insula and the primary motor cortex. The fMRI data was analyzed with the FMRIB Software Library (FSL). We limited the analysis to the DLPFC by filtering the data with a region of interest (ROI) individually defined for each subject based on its own brain anatomy and conservative Talairach coordinates. For the study of fMRI indices, we used the centers of gravity (COG) of activation clusters. The COG is a 3 dimensional coordinate (x, y, z) computed based on the z-values of all voxels constituting a cluster. Results: The paradigm induced activation in the brain areas known to be involved in WM. In response to the WM paradigm, the COGs of the activation clusters in the DLPFC had a significantly greater spatial heterogeneity in patients compared to controls in the left hemisphere. The right hemisphere did not show any significant difference between the two groups. Conclusion: Our hypothesis is supported by our findings in the left hemisphere, but not the right. The methods that were developed for this study are a first attempt to study the spatial heterogeneity of the activation in the DLPFC. The power of the results would benefit from improvement of those methods. In particular, the definition of the DLPFC ROI is to be improved in order to better target the activation patterns of interest