Lesion area assessment in psoriasis patients

Area of Science:

• Dermatology and Medical Imaging
• Computational Pathology
• Image Analysis for Clinical Assessment

Background:

• Psoriasis is a chronic autoimmune skin condition characterized by red, scaly plaques.
• Current treatment efficacy assessment relies on the Psoriasis Area and Severity Index (PASI), often measured manually.
• Manual PASI measurements are subjective, time-consuming, and prone to variability, necessitating objective quantification methods.

Purpose of the Study:

• To develop and validate a computer-based automatic method for measuring the 'area' parameter of psoriasis lesions within the PASI standard.
• To enhance the objectivity and efficiency of psoriasis lesion assessment.
• To improve the detection capabilities for diverse plaque morphologies.

Main Methods:

• Utilized the YCbCr color space to distinguish psoriatic plaques from healthy skin.
• Implemented an optimal thresholding technique for automated plaque segmentation.
• Evaluated the method's performance on clinical image data.

Main Results:

• The automated method achieved high accuracy, exceeding 96% in 18 out of 20 cases and 92% in an additional case.
• Demonstrated superior performance in detecting various plaque types, including those with silvery-white scales, on hairy skin, and tiny lesions.
• Successfully segmented simple, scale-less plaques.

Conclusions:

• The proposed computer-based method offers a precise, automated, and efficient approach to measuring psoriasis lesion area.
• This automated system overcomes limitations of manual assessment, providing consistent and accurate data for PASI scoring.
• The ability to detect diverse plaque types enhances its clinical utility in managing psoriasis.