Validation and clinical use of Doc.Skin, an AI‑Based Skin Type Diagnostic System

Introduction. Skin type classification is a key tool in dermatological and cosmetological practice, providing the foundation for diagnosis and personalized care. The most widely known system is the Leslie Baumann typology, which includes 16 skin types and is commonly used in international practice. However, its application in the Russian population has revealed several limitations: the system does not account for normal and combination skin types while combination skin is predominant among respondents; and it does not incorporate acne proneness, one of the most common dermatological conditions.
Purpose of the study. 1. To assess the diagnostic accuracy and adaptability of the Doc.Skin methodology compared with the Baumann typology in the Russian population. 2. To conduct a clinical validation of the Doc.Skin system by comparing its results with dermatologists’ assessments.
Materials and methods. A single‑center prospective comparative diagnostic study was conducted in two stages. In the first stage, 1,687 respondents completed an online diagnosis using the Doc.Skin system, which is based on artificial intelligence (AI). In the second stage, 250 participants underwent online diagnosis with Doc.Skin immediately prior to in‑person examination. Skin type and its characteristics were verified by board‑certified dermatologists, who were blinded to the system’s outputs. Expert assessments were then compared with the results of Doc.Skin to evaluate its diagnostic accuracy.
Results. Combination skin was the predominant type in the Russian population (43%), followed by normal (24%), dry (18%), and oily (16%). Thus, the Baumann typology, which includes only two basic skin types: dry and oily – covered just 34% of the sample, substantially limiting its diagnostic value in this population. Acne proneness, not represented in the Baumann typology, was identified to varying degrees in 51% of participants and ranked second among prevalent skin concerns. Consequently, the overall adaptability of the Baumann methodology, considering all basic types and skin concerns, was only 17.2 %, compared with the diagnostic coverage of the Doc.Skin system. In the second stage, a high level of concordance was observed between AI‑based Doc.Skin results and dermatologists’ clinical assessments: concordance in skin type and characteristic definitions reached 98.3%.
Conclusions. The AI‑based Doc.Skin diagnostic system, which incorporates 64 skin types and accounts for key characteristics such as sensitivity, acne, wrinkles, and pigmentation, demonstrated high diagnostic accuracy and clinical validity. The methodology can be considered a reliable tool for skin classification in the Russian population and may be applied by dermatologists, healthcare professionals, and in consumer applications to provide personalized skin care recommendations.

1. Ahn S.K. et al. Baumann skin type in the Korean female population // Annals of Dermatology.– 2017. V. 29. № . 5. P. 586–596.

2. Baumann L. The skin type solution: a revolutionary guide to your best skin ever.– Bantam, 2006.

3. Kruglova.L.S. et al. Acne and rosacea. 2021. (In Russ.).

4. Zaenglein A.L. Acne vulgaris //New England Journal of Medicine. 2018. V. 379. № . 14. P. 1343–1352.

5. Østergaard C. E. S. et al. Akne //Ugeskrift for laeger.– 2025.– V. 187.– № . 14.– P. V09240660.

6. Gryazeva N.V. et al. The influence of exposure factors on the course of acne // Medical Alphabet. 2024; (9): 31–37. (In Russ.).

7. Leignadier J. et al. Lysine dendrimer, a new non aggressive solution to rebalance the microbiota of acne prone skin //Pharmaceutics. 202; 5 (8): 2083.

8.