Burn Depth Analysis Using Multidimensional Scaling Applied to Psychophysical Experiment Data

Abstract

In this paper a psychophysical experiment and a multidimensional scaling (MDS) analysis are undergone to determine the physical characteristics that physicians employ to diagnose a burn depth. Subsequently, these characteristics are translated into mathematical features, correlated with these physical characteristics analysis. Finally, a study to verify the ability of these mathematical features to classify burns is performed. In this study, a space with axes correlated with the MDS axes has been developed. 74 images have been represented in this space and a k-nearest neighbor classifier has been used to classify these 74 images. A success rate of 66.2% was obtained when classifying burns into three burn depths and a success rate of 83.8% was obtained when burns were classified as those which needed grafts and those which did not. Additional studies have been performed comparing our system with a principal component analysis and a support vector machine classifier. Results validate the ability of the mathematical features extracted from the psychophysical experiment to classify burns into their depths. In addition, the method has been compared with another state-of-the-art method and the same database.

Features identification for automatic burn classification

“…In other words, each image i, i = 1,Á Á Á,20, is assigned a position in the P-dimensional space, X i = (x 1 ,Á Á Á,x P ), where distances between images are correlated to differences between images according to their similarity indexes. In this experiment P = 3 was identified as the optimum space dimension [21].…”

“…If a colour pixel in an image is expressed in the L*a*b* colour space [38] as I L*a*b* (m,n) = (I L* (m,n), I a* (m,n), I b* (m,n)), where I L*a*b* represents the image in the L*a*b* colour space, the hue average value within a burn can be estimated as [21,38]: and Q represents the set of pixels belonging to the burn. Humidity can be related to the presence of glitters in the photograph.…”

“…Humidity can be related to the presence of glitters in the photograph. Thus, outliers jointly in the red (R), green (G) and blue (B) components of the pixels were potential features to measure this physical characteristic, for glitters tend to be white due to the camera flash [21]. More specifically, if I RGB (m,n)=(I R (m,n), I G (m,n), I B (m,n)) is the RGB image, represents the number of outliers in the RGB components normalized by the area of the burn.…”

“…To measure the colour homogeneity we have chosen the third standardized moments (skewness of a* and b* components: sk a* , sk b* ) and the angular variance of hue, calculated according to refs. [21,27]: where Q* is the set of pixels that, not being considered outliers, belong to the burn and h* is the hue average within the burn except for pixels that are outliers. Colourfulness in Colour Science is traditionally linked to chroma, which is defined as the colourfulness of an area judged as a proportion of the brightness of a similarly illuminated white area.…”

“…Colourfulness in Colour Science is traditionally linked to chroma, which is defined as the colourfulness of an area judged as a proportion of the brightness of a similarly illuminated white area. C à aÃbà is a numerical parameter that tries to measure this colourfulness and its calculation is [21,38]:…”