Robust demarcation of basal cell carcinoma by dependent component analysis-based segmentation of multi-spectral fluorescence images

摘要: This study was designed to demonstrate robust performance of the novel dependent component analysis (DCA)-based approach demarcation basal cell carcinoma (BCC) through unsupervised decomposition red–green–blue (RGB) fluorescent image BCC. Robustness intensity fluctuation is due scale invariance property DCA algorithms, which exploit spectral and spatial diversities between BCC surrounding tissue. Used filtering-based represents an extension independent (ICA) necessary in order account for statistical dependence that induced by similarity generates weak edges what a challenge other segmentation methods as well. By comparative with state-of-the-art such active contours (level set), K-means clustering, non-negative matrix factorization, ICA ratio imaging we experimentally good DCA-based two demanding scenarios where has been varied almost orders magnitude.