AI Model Creates Successful Treatment Regimens for Women with Hair Loss

Image Credit: © Kittiphan - stock.adobe.com

In a recent study, clinicians were able to create personalized treatments for patients with hair loss based on artificial intelligence (AI).¹ The AI platform, developed by MDalgorithms, Inc., was able to recommend therapies that greatly improved hair growth, thickness, fullness, volume, and coverage over 6 months. Hair loss impacts up to 50% of women and 80% of men.²

The prospective, single-center, single-blind clinical trial included 27 women between the ages of 34 and 65 (median age: 60) who had self-reported hair thinning. Most patients had straight or wavy hair. Each participant received a variety of over-the-counter hair loss treatments, which included several combinations of shampoos, topical serums, cosmeceuticals, nutraceuticals, oral supplements, and marine collagen peptides. The products included several well-known active ingredients, including green tea, saw palmetto extract, glycerin, hyaluronic acid, caffeine, ginger extract, biotin, vitamins E, C, and D, zinc, selenium, copper, keratin, and amino acids.

To determine these therapies, an AI analysis of scalp images and patient questionnaires was conducted. Patient photographs were taken via smartphone through the MDhair mobile app. The AI model in this study was previously trained on 47,0000 scalp images and hair loss patterns of all ages, ethnicities, severities, and hair types. The regimens were implemented for 6 months, with evaluations at baseline, week 12, and week 24, along with additional 5- and 7-point self-assessments at weeks 4 and 8. Investigators took digital images and objectively graded the photographs. Hair shed count and transepidermal water loss were also measured.

Patients and clinicians saw significant improvements in hair growth, thickness, fullness, volume, and coverage at all time points (p < 0.001). After 23 weeks, 77.7% of patients had greater hair growth, 62.9% had greater coverage, and 62.9% had more overall thickness. Hair shedding decreased in most patients throughout the study, first by 37.3% at 12 weeks and then by 32.4% at 24 weeks. Scanning electron microscopy also showed that hair texture improved significantly, based on the improvement of cuticle structure and shaft damage.

Transepidermal water loss on the scalp was also reduced by 61.5% at the 12-week mark and 69% at the 24-week mark. Almost 90% of patients observed overall hair improvement. Over 85% experienced better scalp health, while 92.6% said their hair was less brittle (p < 0.001). No treatment-related adverse effects were reported with any of the therapies.

A previous pilot study mirrored these results and found that an AI assessment correctly evaluated 28 out of 30 cases of female androgenetic alopecia, creating an accuracy rate of 94%, demonstrating that these AI systems can perform at the high caliber required of health care providers.³ Although this cannot replace expert clinical advice, the authors hope it can be used in conjunction with in-person treatment, especially with the overwhelming amount of hair loss management therapies available to patients.

This data-driven approach will allow dermatologists and patients to create a customized, cheaper, and more accessible regimen for hair loss management, especially in patient groups who cannot receive in-person dermatological care as often. Future research hopes to combine large language models with this AI technology to further improve recommendations.

"This study highlights the potential of AI-driven personalization in hair care, offering a safe and effective comprehensive inside and outside non-medicated option supported by both subjective and objective clinical outcomes,” study author Yoram Harth, MD, FAAD, told Dermatology Times.

References

1. Bhardwaj V, Rodgers N, Harth O, Harth Y. Artificial Intelligence-Based Personalization of Treatment Regimen for Hair Loss: A 6-Month Clinical Trial. J Drugs Dermatol. 2025;24(3):233-238. doi:10.36849/JDD.8611

2. Piraccini BM, Alessandrini A. Androgenetic alopecia. Ital J Dermatol Venereol. 2014;149(1):15-24.

3. Harth Y, Harth O. Computer vision AI-based androgenetic alopecia analysis using a novel mobile web app. Int Soc Invest Dermatol, Tokyo, May 2023. J Invest Dermatol. 2023;143(5).