20 March 2014 A preliminary study for fully automated quantification of psoriasis severity using image mapping
Author Affiliations +
Abstract
Psoriasis is a common chronic skin disease and it detracts patients’ QoL seriously. Since there is no known permanent cure so far, controlling appropriate disease condition is necessary and therefore quantification of its severity is important. In clinical, psoriasis area and severity index (PASI) is commonly used for abovementioned purpose, however it is often subjective and troublesome. A fully automatic computer-assisted area and severity index (CASI) was proposed to make an objective quantification of skin disease. It investigates the size and density of erythema based on digital image analysis, however it does not consider various inadequate effects caused by different geometrical conditions under clinical follow-up (i.e. variability in direction and distance between camera and patient). In this study, we proposed an image alignment method for clinical images and investigated to quantify the severity of psoriasis under clinical follow-up combined with the idea of CASI. The proposed method finds geometrical same points in patient’s body (ROI) between images with Scale Invariant Feature Transform (SIFT) and performs the Affine transform to map the pixel value to the other. In this study, clinical images from 7 patients with psoriasis lesions on their trunk under clinical follow-up were used. In each series, our image alignment algorithm align images to the geometry of their first image. Our proposed method aligned images appropriately on visual assessment and confirmed that psoriasis areas were properly extracted using the approach of CASI. Although we cannot evaluate PASI and CASI directly due to their different definition of ROI, we confirmed that there is a large correlation between those scores with our image quantification method.
© (2014) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Kazuhiro Mukai, Kazuhiro Mukai, Hitoshi Iyatomi, Hitoshi Iyatomi, } "A preliminary study for fully automated quantification of psoriasis severity using image mapping", Proc. SPIE 9035, Medical Imaging 2014: Computer-Aided Diagnosis, 90353N (20 March 2014); doi: 10.1117/12.2043446; https://doi.org/10.1117/12.2043446
PROCEEDINGS
6 PAGES


SHARE
Back to Top