Texture features for classification of skin scar multi-photon fluorescence microscopic images

We implemented a method that combined local binary pattern (LBP) operator and wavelet transform to extract texture features of multi-photon fluorescence microscopic images from human physiologic and pathologic scars. The proposed method includes texture feature extraction in scar multi-photon fluorescence microscopic images, and classification with a support vector machine. By comparing with other texture feature extraction methods with respect to the accuracy and receiver operating characteristic analysis, the method combining LBP operator and wavelet transform was demonstrated to achieve higher accuracy and better performance. It can provide a basis for improving clinical diagnosis of scar types and lead to the development of new therapeutic methods for dermatology and plastic surgery.