Quantitative analysis of pulmonary structures in PMCT; Stereological comparison of drowning compared to opioid-overdose cases.

Objectives: The aim of this study, was to introduce stereology as a versatile and robust tool for quantitative image analysis of volume and attenuation characteristics (Hounsfield Units (HU's)), in a blinded control case study investigating lungs of drowning victims compared to a control group on post mortem computed tomography (PMCT) data. Materials and Methods: PMCT scans of the lungs from 14 drowned cases and 14 matched opioid-overdose-controls was included. Quantitative CT-analysis was performed using a stereological approach adapted to PMCT data that allowed for precise extraction of volume and HU-values using stereological point-probes assigned manually to individual lung-structures. Qualitative radiological image interpretation performed by a trained radiologist was compared to the quantitative analyses. Results: No significant difference was found for total lung volume, volumes of consolidations, ground glass opacities, bronchi, and air-filled lung tissue. When comparing drowning cases with opioid overdose cases as controls, the extracted HU-values did not show statistically significant changes in mean attenuation characteristics. No major discrepancies were found between the quantitative analysis and qualitative analysis. Conclusion: Conventional radiological evaluation of PMCT images rely on the radiologists’ ability to distinguish normal from pathological. Quantitative image analysis offers, to name a few, precise estimations of structure volume and HU-statistics. Although the used 14 matched cases data failed to significantly aid the diagnosis of drowning statistically, we envisage that quantitative PMCT analysis using stereology could become a valuable tool to improve objective forensic radiological interpretation.