MORPHOMETRIC IMAGE ANALYSIS COMPARISON OF BASAL CELL CARCINOMAS & MELANOMAS OF THE HEAD & NECK VERSUS OTHER SITES

Abstract

AbstractBackgroundManagement of skin cancers are heavily dependent on clinical diagnosis via dermoscopy. Dermoscopic morphology of basal cell carcinomas (BCC) and melanomas influence diagnosis and may be influenced by anatomic location.ObjectiveIn this study, we aimed to investigate the morphologic differences between anatomic sites of the head and neck (H&N) versus other body sites in basal cell carcinomas and melanomas.Patients and methodsThe publicly available HAM10000 dataset was used in this study. Morphometric image analysis of the BCCs (n=422) and melanomas (n=868) in this dataset was performed using an open-source image analysis software. Univariate and multivariate statistical analysis was done to identify differences between H&N and other anatomic sites. The multifactorial data was further interrogated with dimensionality reduction techniques, linear discriminant analysis, principal component analysis and t-distributed stochastic neighbour embedding.ResultsUnivariate and multivariate statistical analyses revealed significant differences between H&N and other sites for both BCCs and melanomas (P<0.05). Fifty-three univariate and 11 multivariate features were found to be statistically significant in the BCC group. Thirteen univariate and 8 multivariate features were statistically significant in the melanoma group. Dimensionality reduction via linear discriminant analysis of the BCC groups revealed modest separation of the data by anatomical site. However, melanomas appeared to be more homogenous across H&N and other body sites.ConclusionBCCs of the H&N may be morphologically different to BCCs of other body sites. This may influence the accuracy of computer-assisted diagnostic tools and specialist clinicians working predominantly in the H&N region should exercise caution when employing these tools in clinical practice.