Simple horizontal averaging programme enables shade correction for image analysis in psoriasis

Computer-assisted area measurement of skin lesions would be useful for the assessment of disease severity. Difficulties can be encountered, however, when lesions cannot simply be separated using the normal threshold tool of an image analyser. Owing to body curvature, peripheral sites are usually shadowed. To obtain a well-fitting curved line to account for body curvature-induced glare and shadow, we devised a simple C-programme for averaging neighbouring pixels on the same horizontal line. An image was captured from a colour slide to the image analysis system. C++ was used for programming. Execution of the averaging programme with the original image resulted in a background brightness image file. When this was subtracted from the original image, the lesions were easily detected using the standard threshold tool and the percentage of area involved was calculated. This method is sufficiently accurate for the assessment of disease severity.