Dermaself, a Milan-based beauty tech company, is integrating its AI-powered skin analysis technology into remote clinical studies. This innovation allows testing laboratories to conduct efficacy trials with volunteers capturing their own photos at home.

The AI, which measures over 60 visible skin parameters from a single selfie and recommends personalized routines based on product INCI lists, is now being utilized to track skin changes over time in controlled studies. The technology’s foundation stems from Derming, an independent research institute in Monza with over 30 years of dermatological expertise. They validated the skin taxonomy behind the AI using more than 100,000 clinical images reviewed by board-certified dermatologists.

Remote Studies Replace Traditional Site Visits

Dermaself developed a specialized version of its platform for testing laboratories. This platform enables studies where volunteers take standardized photos at home, eliminating the need for travel to a test site. The core advantage lies in measurement consistency. The AI must accurately assess skin under varying lighting conditions from any smartphone, a challenge akin to comparing a photo taken in week eight to a baseline image.

The platform employs patent-pending color normalization to adjust for lighting and color variations in every image. On repeated measurements of the same skin area, the AI achieves a repeatability score of 0.92–0.93, making the transition from routine recommendation to study measurement seamless.

A typical remote study follows a five-step process: setup, baseline capture, checkpoint captures, automatic comparison, and reporting. Laboratories define the protocol and endpoints, selecting from over 60 skin parameters grouped into nine categories. A study can be launched within five days.

Volunteers receive on-screen guidance for taking standardized photos at home or on-site. At each checkpoint—usually at two, four, and eight weeks—the AI measures every image using consistent criteria and compares results to each volunteer’s baseline without manual intervention.

Laboratories receive individual and cohort-wide results, with their experts drafting the final report and conclusions. This method eliminates two major constraints in traditional efficacy testing: limited test site capacity and the requirement for volunteers to travel frequently.

Read Also: UK shoppers still shun AI product recommendations

Remote studies can accommodate up to 10,000 volunteers, allowing cohort sizes to align with research needs rather than physical limitations. Participants apply products and photograph their skin in real-life conditions, providing more practical usage insights.

Dermaself employs patent-pending 3D anonymization to permanently remove biometric identity from volunteer faces while preserving the skin data necessary for analysis. Data protection terms are established with each laboratory prior to the commencement of studies.

Study Design Flexibility and Capture Settings

Dermaself’s technology applies a single measurement across three capture settings: at home, in a test center, and in the lab. It corrects light and color variations in every image. Laboratories can design studies based on volunteer locations, even mixing capture locations within a single protocol.

A common approach involves taking a baseline measurement in the lab, followed by checkpoints photographed by volunteers at home. This hybrid method allows labs to validate initial measurements under controlled conditions while still benefiting from remote follow-ups.

Remote studies on the Dermaself platform can range from single product tests with a few checkpoints to long-term tracking over several months. Since volunteers photograph their skin at home rather than traveling to a testing site for each visit, long-term follow-ups are more manageable for both volunteers and laboratories.

The AI does not replace expert graders. It automates the repetitive scoring of study images, while study design, expert evaluation, and conclusions remain the responsibility of the laboratory. The decision on whether the data supports a specific product claim rests with the lab and the brand’s regulatory team.