Psoriasis segmentation through chromatic regions and Geometric Active Contours

Area of Science:

• Dermatology
• Medical Image Analysis
• Computer Vision

Background:

• Psoriasis is a chronic skin condition requiring accurate lesion measurement.
• Manual segmentation of numerous psoriasis lesions in digital images is time-consuming and subjective.

Purpose of the Study:

• To develop and validate an automated algorithm for segmenting psoriasis lesions in full-body digital photographs.
• To compare the algorithm's segmentation performance against human operators.

Main Methods:

• A novel algorithm combining chromatic information for initial lesion zone identification.
• Utilizing Geometric Active Contours for precise plaque segmentation.
• Evaluation of segmentation variability against multiple human operators.

Main Results:

• The algorithm successfully isolates and segments psoriasis plaques from full-body images.
• The segmentation variability of the algorithm is comparable to the variability observed between human operators.
• The approach is effective even with a large number of discrete lesions.

Conclusions:

• The proposed automated method offers a reliable and consistent approach for psoriasis lesion segmentation.
• This algorithm has the potential to improve the efficiency and objectivity of psoriasis assessment in clinical and research settings.
• The developed technique demonstrates performance on par with human expert segmentation.