Detection and volume estimation of artificial hematomas in the subcutaneous fatty tissue: comparison of different MR sequences at 3.0 T

Comparison of visible-light and infrared photography for visualizing hematomas postmortem

Photography is essential in forensic medicine documentation. While visible-light photography uses the human eye's spectrum (approximately 380-780 nm), infrared (IR) photography captures wavelengths invisible to the naked eye (approximately 700-1100 nm). This study aimed to assess the reliability of IR photography in detecting subcutaneous hematomas in deceased individuals. In postmortem examinations of 23 individuals with different skin tones, 43 hematomas were evaluated; for ethical reasons, hematomas on the face, neck, hands, or feet were excluded. Standardized photographs were taken using two different cameras: a Nikon D810 (visible-light) and a Nikon D800E modified with a 700 nm IR filter. Subsequently, tissue samples including the hematomas were excised. Hematoma density was assessed on paraffin-embedded samples using a Keyence VHX 5000 digital microscope. Raw IR photographs were processed with Photoshop to obtain tonal values of the darkest hematoma spot and the brightest spot of the surrounding intact tissue. Visual inspection of the excised samples confirmed that infrared photography accurately depicted 100 % of the 43 hematomas, whereas using visible-light photography, only 53.5 % were well visible and 46.5 % poorly visible. Tonal values correlated positively with microscopic densities of the hematomas, yielding a moderate to strong linear correlation coefficient of 0.70 (p < 0.001). In conclusion, IR photography is highly reliable in visualizing subcutaneous hematomas and has clear advantages over visible-light photography. Our results suggest that IR photography could be valuable as an additional tool in depicting suspected hematomas in living individuals.

Extrapolation study for determining the time since injury in a rat subcutaneous hematoma model utilizing ATR-FTIR spectroscopy

The determination of the time of an injury has been a major problem in forensic science due to the lack of objective, reliable and portable methods. In this study, a subcutaneous hemorrhage model in rats was established over six days, and attenuated total reflection-Fourier transform infrared (ATR-FTIR) spectroscopy coupled with chemometrics was used to determine the time since injury. Initial principal component analysis (PCA) showed variance among hematoma sites. Subsequently, spectral data were acquired to establish a dependable partial least square (PLS) regression model with predictive abilities. The root mean square error of cross-validation (RMSECV) and the root mean square error of prediction (RMSEP) values produced by a genetic algorithm (GA) were 0.64 d (R2 = 0.88) and 0.57 d (R2 = 0.90), respectively. Few variables were involved in the model, and significantly better results were obtained in comparison to the conventional full-spectrum PLS model. In combination with the results of variable importance in projection (VIP) scores, all components, including proteins, nucleic acids and phospholipids, provided inferences regarding the samples at different time points; additionally, amide I and II bands represented the secondary structure of proteins and provided the largest contribution. Based on our preliminary study, the combination of swift and nondamaging ATR-FTIR spectroscopy with chemometrics could prove to be an advantageous approach for gauging the age of an injury in the forensic field.