Artificial Intelligence-Driven Skin Parameter Assessment in Indian Adults: Establishing Population-Specific Reference...

Background Artificial intelligence (AI)-based skin analysis enables objective, large-scale assessment of dermatologic features; however, population-specific normative reference ranges for Indian skin remain limited. This exploratory study aimed to establish percentile-based reference distributions for facial skin parameters in Indian adults, predominantly comprising Fitzpatrick skin types III-V, who demonstrate distinct pigmentary and aging patterns that are underrepresented in the existing literature. Methods This retrospective, cross-sectional study analyzed de-identified facial scan data from 20,914 healthy Indian adults (≥18 years) obtained through the SkinSensAI™ platform (Ningen Skin Sciences Pvt. Ltd., New Delhi, India), of whom 20,913 with complete age data were included in age-stratified analyses. Facial images were processed using an AI engine to generate standardized severity scores (1-5 scale) for 12 facial skin parameters. Descriptive statistics, age-, gender-, and Fitzpatrick-stratified analyses, and inter-parameter correlations were performed in accordance with the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines. Results Dark circles and enlarged pores emerged as the most prevalent concerns, while wrinkle severity remained low in participants younger than 45 years. Wrinkles showed the strongest correlation with age (r = 0.57, p < 0.001). Male participants exhibited significantly higher scores for pores, spots, wrinkles, and skin laxity than female participants (p < 0.001). Darker Fitzpatrick skin types demonstrated a higher pigmentary burden, whereas wrinkle severity was relatively lower, consistent with melanin-associated photoprotection. Distinct correlation clusters reflected aging-related, pigmentation-sebaceous, and independent skin domains. Conclusion This study establishes the first large-scale, AI-derived reference database of facial skin parameters specific to the Indian population. These population-specific benchmarks support objective dermatologic assessment, personalized patient counseling, and equitable AI application in clinical dermatology.