Simple Burn Wound Severity Assessment Classifier Based On Spatial Frequency Domain Imaging (SFDI) and Machine Learning

摘要

Assessment of burn severity is critical for wound treatment. Spatial frequency domain imaging (SFDI) has beenpreviously used to characterize burns based on the relationships between histology and tissue optical properties. Recently,multispectral and hyperspectral imaging optical features have been combined with machine learning to classify burnseverity. Here, we investigated the use of SFDI reflectance data at multiple wavelengths and spatial frequencies, with asupport vector machine (SVM), to predict severity in a porcine model of graded burns. Burn severity predictions usingSVM were compared to burn grade determined using histology techniques. Results suggest that the combination of spatialfrequency data with machine learning models has the potential for accurately predicting burn severity at the 24 hr postburntime point.