Parametric Acne Severity (PAS) Score and Lesion Counts From Multi-Modality Facial Image Analysis Correlates Strongly...

Area of Science:

• Dermatology
• Medical Imaging
• Computational Analysis

Background:

• Acne clinical trials face challenges due to subjective lesion counting and severity assessments, leading to increased costs and hindering meta-analyses.
• Standardized, objective outcome measures are crucial for reliable acne treatment efficacy comparisons.

Purpose of the Study:

• To evaluate objective measures for acne lesion counts and global severity using multimodal photography analysis.
• To develop and validate an algorithm for acne lesion counting and a Parametric Acne Severity (PAS) score.

Main Methods:

• An algorithm was developed and validated for counting inflammatory and non-inflammatory acne lesions in 30 subjects.
• A composite score (PAS) was generated using photographic data on lesion topography, erythema, and C. acnes fluorescence.
• Algorithm-based lesion counts and PAS scores were compared against expert rater assessments and Investigator's Global Assessment (IGA).

Main Results:

• The lesion counting algorithm demonstrated strong correlations with expert raters (0.77 for inflammatory, 0.85 for non-inflammatory lesions).
• The Parametric Acne Severity (PAS) score showed a significant correlation with rater IGA grades (0.82).
• No relationship was found between lesion counts and IGA, highlighting the limitations of traditional global assessments.

Conclusions:

• Objective photographic measures, including lesion counting algorithms and the PAS score, are reliable and correlate well with expert clinical judgment.
• Implementing these objective measures in acne clinical trials can reduce variability, standardize outcomes, and enhance the analysis of treatment effects.