Post-mortem MRI-based volumetry of the hippocampus in forensic cases of decedents with severe mental illness.

Abstract

A decrease in the volume of the hippocampus is associated with severe mental illness, especially schizophrenia, and has been studied extensively in the living using magnetic resonance imaging. Autopsy cohorts also represent a valuable data source for imaging studies. However, post-mortem magnetic resonance imaging (PMMRI) is subject to unique challenges, such as the lower core temperature of scanned subjects and the influence of decomposition processes. This study aimed to determine if results from in vivo studies could be replicated on a post-mortem cohort of decedents who suffered from severe mental illness. We included 96 decedents with either schizophrenia (n = 34), depressive disorder (n = 17), or no known psychiatric diagnosis (n = 45) from April 2015 to January 2017. All cases underwent a T2-weighted cerebral MRI less than 24h before autopsy. We used a manual segmentation algorithm to define the hippocampus on coronal images and subsequently estimate the volume of the region. The group with schizophrenia had a statistically significant 9.5% decrease in mean hippocampal volume compared with control subjects, while the group with depression trended towards a reduced volume, but this difference was not statistically significant. Thus we were able to replicate previous results from in vivo studies. PMMRI has unique potential for research in that it can be combined with procedures possible only in the research fields of clinical pathology and forensic science, e.g. histopathological sampling.