Characterization and pattern recognition of color images of dermatological ulcers - a pilot study.

We present color image processing methods for the characterization of images of dermatological lesions for the purpose of content-based image retrieval (CBIR) and computer-aided diagnosis. The intended application is to segment the images and perform classification and analysis of the tissue composition of skin lesions or ulcers, in terms of granulation (red), fibrin (yellow), necrotic (black), callous (white), and mixed tissue composition. The images were analyzed and classified by an expert dermatologist following the red-yellow-black-white model. Automatic segmentation was performed by means of clustering using Gaussian mixture modeling, and its performance was evaluated by deriving the Jaccard coefficient between the automatically and manually segmented images. Statistical texture features were derived from cooccurrence matrices of RGB, HSI, L∗a∗b∗, and L∗u∗v∗ color components. A retrieval engine was implemented using the knearest-neighbor classifier and the Euclidean, Manhattan, and Chebyshev distance metrics. Classification was performed by c ©2014 by Lucas C. Pereyra, Silvio M. Pereira, Juliana P. Souza, Marco A. C. Frade, Rangaraj M. Rangayyan, Paulo M. Azevedo-Marques. ∗This work was partially supported by The National Council for Scientific and Technological Development (CNPq) — grants 472508/2010-5, 304225/2010-0, and 573714/2008-8 (INCT/INCoD), and the Natural Sciences and Engineering Research