Can intramuscular hemorrhage in the back be detected on postmortem CT? A forensic study in drowning cases

An intramuscular hemorrhage in the back is considered a potential supportive finding with regard to diagnosing drowning; however, its detection traditionally requires autopsy. This study aimed to evaluate whether back muscle intramuscular hemorrhage can be reliably identified on postmortem computed tomography (PMCT). Here, we retrospectively reviewed 31 drowning cases for which autopsies had already been performed, comparing PMCT scans acquired before dissection with the subsequent autopsy findings. Intramuscular hemorrhage in 11 major back muscles was recorded and compared with the PMCT findings; furthermore, diagnostic accuracy (sensitivity, specificity, and F1 score) was calculated. Of the 682 muscles, 176 (25.8 %) showed hemorrhage on autopsy. Overall, the PMCT sensitivity was 21.6 %, specificity was 92.1 %, and F1 score was 0.299. ROC analysis showed that hemorrhages larger than certain size thresholds (6.0 and 4.5 cm in vertical and horizontal dimensions, respectively, with 16.0 cm² area) were more likely to be detected. Detection improved with the increase in hemorrhage size; however, many smaller lesions were missed. While large hemorrhages can occasionally be recognized on PMCT, routine forensic applications for detecting back muscle hemorrhages remain limited. PMCT often underestimates the incidence of smaller intramuscular hemorrhages. Meticulous autopsy remains indispensable for the accurate diagnosis of intramuscular hemorrhage in drowning cases.