Injuries of the isolated larynx-hyoid complex in post-mortem computed tomography (PMCT) and post-mortem fine preparation (PMFP). A comparison of 54 forensic cases

To assess the diagnostic accuracy (ACC) of post-mortem computed tomography (PMCT) for fractures of the isolated larynx-hyoid complex (LHC) in comparison to post-mortem fine preparation (PMFP). This monocentric prospective study enclosed 54 LHCs that were extracted during autopsy, fixed in formalin, and underwent a PMCT scan (64-row multidetector CT, helical pitch). Two radiologists independently analyzed the LHC scans for image quality (IQ) and fractures (4-point Likert scales). A specialized forensic preparator dissected the specimens under the stereomicroscope. The PMFP results were standardized documented, and used as the standard of reference for the comparison to PMCT. The PMCT-IQ of 95% of the LHC images was rated as good or excellent. IQ was decreased by decay, incisions during autopsy, and separation of the hyoid from the cartilaginous components in 7, 3, and 12 specimens, respectively. PMFP detected 119 fractures in 34 LHCs (63.0%). PMCT identified 91 fractures in 32 specimens (59.3%). PMFP and PMCT significantly agreed concerning the location (Cohen's κ = 0.762; p < 0.001) and the degree of dislocation (κ = 0.689; p < 0.001) of the fractures. Comparing PMCT to PMFP resulted in a sensitivity of 88.2%, a specificity of 90.0%, and an ACC of 88.9% for the LHC. The ACCs for the hyoid, thyroid, and cricoid were 94.4%, 87.0%, and 81.5%, respectively. PMCT procedure was significantly faster than PMFP (28.9 ± 4.1 min vs. 208.2 ± 32.5 min; p < 0,001). PMCT can detect distinct injuries of the isolated LHC and may promptly confirm violence against the neck as cause of death. PMFP outmatches PMCT in the detection of decent injuries like tears of the cricoid cartilage.