Fig. 2: Contribution of regional cortical thickness measures to the person-based similarity index.
From: Personalized estimates of morphometric similarity in bipolar disorder and schizophrenia
In each participant within each diagnostic group, we used the leave-one-out approach to re-compute the person-based-similarity index for cortical thickness (PBSI-CT) after removing one regional cortical thickness measure at the time. We then calculated the absolute difference between each recalculated PBSI-CT and the original PBSI-CT (i.e., which included all cortical regions). The absolute mean of these difference scores in each diagnostic group is shown mapped on the cortical surface; warm colors reflect higher regional contributions. There were no significant differences in regional contributions between diagnostic groups.
