The acne artificial intelligence review finds diagnosis tools dominate research but rarely test results across skin tones. That gap leaves equity performance largely untested for teledermatology patients, meaning care through photos or video.
The International Journal of Dermatology reports the review covered 105 articles, with 96.2% on diagnosis only and 76.2% using image-based deep learning the review page. Most tools use deep learning, a pattern-finding method trained on photos, to classify lesions rather than guide treatment. That focus makes image diversity central to safety.
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 studies test?
- Why does skin tone change a photo diagnosis?
- Does assisted diagnosis fix the gap?
- How to make a teledermatology visit safer?
What did the studies test?
Most studies built photo classifiers for diagnosis. In the review counts, 9.5% covered management only and 5.7% covered both diagnosis and management. Treatment support was rare.
Only 13% of the 105 articles reported patient skin color or ethnicity, according to the U.S. National Library of Medicine PubMed record the PubMed record. Only four studies included a full spectrum of diverse skin tones. Equity testing was therefore narrow.
Why does skin tone change a photo diagnosis?
In darker skin, redness looks less obvious and dark spots last longer after breakouts. The Medscape interview with Jules B. Lipoff MD notes erythema cues differ and pigment changes dominate.
Poor light can hide those cues further. Nature Medicine describes a 364-image test across 46 diseases where dermatologists reached 38% top-3 accuracy and generalists reached 19%. Top-3 accuracy means the correct disease was among three guesses. Both groups scored 4 points lower on dark Fitzpatrick V-VI skin, meaning darkest brown skin, than on light skin.
Does assisted diagnosis fix the gap?
Northwestern University reports fair decision support raised accuracy by over 33% overall, about 33% for dermatologists and 69% for primary-care doctors the Northwestern summary. Support means the tool suggests options while the clinician decides. Gains were real but uneven.
That same program widened the light-versus-dark gap for generalists. Unreliable skin-tone labels hide such gaps, according to the npj Digital Medicine scale study. Consumer lesion apps also showed poor sensitivity and specificity, meaning true-positive and true-negative rates, on diverse skin in UCLA testing.
How to make a teledermatology visit safer?
Treat any acne tool as screening, not diagnosis. The Medscape interview with Jules B. Lipoff MD advises using AI only as screening with clinician oversight the dermatology interview.
A clinician should confirm type, severity, and treatment. Bring prior photos to show change over time. Ask the clinic how its photo tool was tested on darker skin before you act on its score.
- Send sharp, well-lit photos without flash glare or heavy filters.
- Include front, angle, and close views plus lesion history.
- List current products, prescriptions, and pigment-change concerns.
You Might Also Like
- What 105 Studies Reveal About Acne AI in the 2026 Review: Accuracy and Validation Questions for Teledermatology Patients
- Skin-Tone Equity Gaps in 2026 Acne AI Research: What the Review Means for Parents Using Acne Apps for Teens
- Skin-Tone Equity Gaps in 2026 Acne AI Research: What Developers Still Need to Prove for People With Skin of Color