Technical note: derivation and evaluation of practical PMCT display window parameters for frozen brain tissue

Frozen cadavers are common in forensic practice. When examining frozen cadaver brains with conventional clinical CT brain windows, brain tissue exhibits a "completely black" appearance precluding assessment. This study aimed to derive and evaluate CT display window parameters for frozen brain tissue to optimize image quality and assist forensic image interpretation.

A two-phase study was conducted on 30 frozen cadavers (-20 °C, ≥ 72 h). In Phase I (n = 2), Window Level (WL) and Window Width (WW) were derived from CT values of anatomical regions on PMCT and locked. In Phase II (n = 28), parameters were applied to a held-out internal evaluation cohort. Effectiveness was evaluated using display coverage rate and subjective scores by two independent raters.

Brain density dropped to negative HU upon freezing. The derived parameters (WL: -10 HU, WW: 110 HU) markedly improved the "completely black" appearance, achieving display coverage rates of 100% for normal brain ROIs and 54.55% for hemorrhagic ROIs, with an overall rate of 96.69%. Subjective image quality scores were significantly higher for the derived parameters (median [IQR]: 18.0 [17.0-18.0]) than for conventional clinical windows (median [IQR]: 6.5 [6.0-7.0]) for both raters (P < 0.001), with excellent inter-rater agreement (overall weighted κ = 0.981).

This study derived and internally evaluated practical display window parameters (WL: -10 HU, WW: 110 HU) for frozen brain PMCT. These parameters effectively improved visualization of frozen brain anatomy and hemorrhagic lesions, providing a pragmatic post-processing display parameters for post-mortem cranial imaging evaluation of frozen cadavers in forensic PMCT practice.