Diagnostic accuracy of unenhanced post-mortem CT and MRI compared to the non-forensic clinical autopsy: a prospective blinded study

Abstract The last decades have seen a constant decline in non-forensic clinical autopsy rates worldwide. In this context, post-mortem computed tomography (PMCT) and post-mortem magnetic resonance imaging (PMMR) might offer an alternative to the clinical autopsy. So far, post-mortem imaging is used routinely only in forensic medicine, but it has not yet been implemented into routine clinical pathology casework. This study aimed to assess the diagnostic accuracy of unenhanced PMCT and PMMR for basic pathology groups and specific diagnoses compared to the clinical autopsy. Post-mortem imaging (PMCT and PMMR) was conducted before autopsy on n = 120 non-forensic patients deceased in hospitals in a prospective study. Imaging findings were compared to autopsy findings, and sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) were calculated for causes of death and specific pathologic findings. Some of the diagnoses recorded showed high specificity and NPV of over 85% for both PMCT and PMMR. However, even the combined use of PMCT and PMMR could not visualize various relevant autopsy findings such as those related to sepsis, hematologic malignancies, chronic liver congestion, endomyocardial, myocardial, and liver fibrosis, acute tubular necrosis, aortic valve stenosis, duodenal ulceration, and small macroscopic findings in general. For specific findings, post-mortem imaging showed no significant differences to autopsy and high diagnostic accuracy with over 85% sensitivity. Examples for such findings included the diagnosis of acute myocardial infarction and pulmonary embolism in PMMR, pneumonia in PMCT, as well as pancreatitis, abscesses, metastases, and aortic dissection in both PMCT and PMMR.