Medical Imaging using Automatic Region of Interest Segmentation for Psoriasis Diagnosis

Medical imaging technology plays an important role for the current healthcare diagnosis and application. Psoriasis is a common skin disease characterized by red plaques which are covered with silvery scales. Psoriasis Area and Severity Index (PASI) is currently the standard to assess the treatment efficacy by evaluating the area and other features. In this paper, we proposed an automatic region of interest segmentation algorithm that can objectively measure the PASI area score. The color space analysis is applied based on YC_bC_r and YIQ. K-mean clustering is used to separate the background from the skin region and segment the region of interest (ROI) from the normal skin. Ten skin image samples from the hospital experimental database have been analyzed in the experiments. The results presented that the proposed algorithm is effective to segment the psoriasis area and measure the region of interest with practical precision for psoriasis diagnosis.