Body surface area measurement and soft clustering for PASI area assessment

Area of Science:

• Dermatology
• Medical Imaging
• Computational Analysis

Background:

• Psoriasis is a prevalent chronic skin condition, with plaque psoriasis being the most common form.
• Current assessment methods like the Psoriasis Area and Severity Index (PASI) suffer from inter- and intra-rater variability.
• Objective and reliable assessment is crucial for effective long-term psoriasis management.

Purpose of the Study:

• To develop and validate an enhanced imaging and analysis system (α-PASI) for objective psoriasis area scoring.
• To improve the accuracy of body surface area (BSA) and lesion area determination in psoriasis patients.
• To assess the reliability and agreement of the α-PASI system compared to traditional methods.

Main Methods:

• Development of enhanced imaging techniques for BSA and lesion area calculation.
• Validation of the BSA determination method using a medical mannequin across four body regions.
• Application of fuzzy c-means clustering for lesion area analysis.
• Double assessment of 46 patient images using the α-PASI area algorithm.

Main Results:

• Validated BSA determination achieved high accuracies: 97.80% (lower limb), 92.41% (trunk), 87.72% (upper limb), and 83.82% (head).
• The α-PASI system demonstrated substantial agreement with kappa coefficients ≥ 0.72 for all body regions (overall kappa = 0.80).
• High reliability was observed for the α-PASI area system in psoriasis area assessment.

Conclusions:

• The developed α-PASI system offers a reliable and objective method for psoriasis area assessment.
• Enhanced imaging and fuzzy c-means clustering significantly improve the accuracy of BSA and lesion area determination.
• The α-PASI system shows potential for consistent and reproducible psoriasis severity scoring in clinical practice.