The March 2026 scoping review of acne AI shows the tools focus almost entirely on diagnosis from images. For teledermatology patients, that means AI may help grade your photos while a dermatologist still decides treatment. A scoping review maps published research on a topic. This one covered diagnosis and management of acne across 105 studies.
Medical information disclaimer: This article is for general educational purposes only and does not provide medical advice, diagnosis, or treatment. Always consult a physician or other qualified health professional about symptoms, medications, tests, or treatment decisions.
Table of Contents
- What did the review examine?
- How do these tools work in a photo visit?
- How accurate is acne grading in practice?
- What limits should patients know?
What did the review examine?
Frederickson and colleagues published their review in the International Journal of Dermatology after online release in November 2025. The team searched PubMed, Cochrane and Scopus for acne plus AI terms and included 105 articles, according to the journal review. Diagnosis dominated the set.
About 96.2% of articles addressed diagnosis only, while 9.5% addressed management only and 5.7% addressed both. That split matters for remote care. Photo-based visits depend on visual review, so diagnostic models fit that workflow more directly than treatment-planning tools.
How do these tools work in a photo visit?
Most studied models read still images. Deep learning appeared in 76.2% of studies, ensemble models in 11.4% and classical machine learning in 9.5%.
Mean accuracy was highest for ensembles at 89.7%, followed by deep learning at 88.5%. Teledermatology often uses store-and-forward photos, live video or hybrid photo-plus-video care. You can make a photo review easier with clear inputs:.
- Use daylight or bright even room light without flash glare
- Send front, left-side and right-side views of the same area
- Include one close view that shows active spots clearly
- Note new products, prescriptions and symptom changes
How accurate is acne grading in practice?
A Japanese study of standardized facial images reported 90.0% overall accuracy for automated acne-severity classification. It also reported a 0.885 macro-average F1-score, a balanced measure of precision and recall, according to the peer-reviewed study. That result shows what a controlled, single-population model can cite.
Lighting, cameras, makeup, acne type and skin tone can change performance outside that setting. Ask your clinic what its tool was tested on. A number from one population does not guarantee the same result for your skin and photos.
What limits should patients know?
Performance drops on darker skin when training images skew light. That bias prompted December 2024 FDA guidance asking developers to validate AI devices across ethnicity, gender, severity and care settings, as summarized by Atta Systems reporting on the FDA guidance.
Acne AI also remains assistive rather than standalone. The only FDA-cleared dermatology AI app as of 2024 was DermaSensor, cleared in January 2024 for physicians checking lesions suggestive of melanoma and other skin cancers in patients 40 and older, not acne. Before a remote visit, ask who reviews the AI result, how your photos are stored and what happens if the grade looks wrong for your skin tone.
You Might Also Like
- What 105 Studies Reveal About Acne AI in the 2026 Review: What Developers Still Need to Prove for People Using Acne Apps
- The 2026 Global Delphi Consensus on Acne and the Microbiome: What the Research Means for Adults With Persistent Acne
- Skin-Tone Equity Gaps in 2026 Acne AI Research: Why Clinician Review Still Matters for People With Skin of Color