AI Accurately Scores Eczema Severity Via Smartphone Photos

A groundbreaking artificial intelligence tool is set to change how atopic dermatitis, commonly known as eczema, is assessed, offering a bridge between patient self-reporting and clinical evaluations. This innovative model, detailed in a study published in Allergy, uses smartphone photographs to objectively measure eczema severity, demonstrating a strong alignment with assessments made by dermatologists.

How the AI Model Works Its Magic

Developed by a collaborative team from Keio University School of Medicine, Kyoto Prefectural University of Medicine, Teikyo University, and Atopiyo LLC, this AI model was trained using an extensive dataset. Over 57,000 images and symptom comments from more than 28,000 users of Atopiyo, Japan's largest digital platform for atopic dermatitis, formed the foundation of its learning. The AI integrates three key components: an algorithm for detecting body parts, a system for identifying lesions, and an engine for scoring severity based on the Three Item Severity (TIS) scale. The TIS scale evaluates erythema (redness), edema/papulation (swelling/bumps), and excoriation (scratch marks) on a 0–9 scale.

Takeya Adachi, MD, PhD Photo courtesy of LinkedIn

Empowering Patients Through Objective Tracking

Dr. Takeya Adachi, the study’s corresponding author from Keio University School of Medicine, highlighted the practical benefits for patients. "Many patients with eczema struggle to evaluate their disease severity on their own," Dr. Adachi stated in a press release. "Our AI model allows for objective, real-time tracking using just a smartphone, empowering patients and potentially improving disease management." This accessibility could be a game-changer for individuals managing this chronic condition.

Putting the AI to the Test Validation and Accuracy

To develop the AI-based TIS (AI-TIS), Adachi and his colleagues utilized Single Shot Multibox Detector and convolutional neural networks for identifying lesions and body parts. They employed a stacking ensemble technique to derive the AI-TIS scores. The model's effectiveness was validated by comparing its scores against TIS scores assigned by dermatologists, as well as the SCORAD and objective-SCORAD indices, which are standard clinical measures.

In a validation set comprising 220 images, the AI-TIS scores demonstrated a strong correlation with clinician-determined TIS scores (R = 0.73; P < .001). Further comparisons involving a subset of 15 patients revealed significant correlations: R = 0.61 with TIS (P = .01), R = 0.53 with objective-SCORAD (P = .04), and R = 0.4 with total SCORAD (P = .12). Impressively, the AI tool achieved 98% accuracy in detecting body parts and 100% accuracy in identifying eczema lesions.

The Gap Between Visible Severity and Patient Experience

An interesting finding was the weak correlation between AI-TIS scores and patient-reported itch, measured using the 0–5 Itch-NRS scale (R = 0.11; P < .001). This observation aligns with clinical experience, where the subjective intensity of pruritus (itch) does not always match the visible severity of the skin condition. This underscores the importance of incorporating objective digital biomarkers in the management of chronic dermatological diseases, providing a more complete picture beyond subjective symptoms.

Who Benefited and The Path Ahead for Teledermatology

The study included 900 participants with a median age of 33 years (ranging from 2 to 71), with women constituting 68% of the group. The median duration of the disease among participants was 25 years. The researchers emphasized the model's significant potential to support remote monitoring of patients, particularly in primary care settings and within the growing field of teledermatology.

The Future of AI in Dermatological Care

Concluding their findings, the investigators stated, "The AI model developed in this study has the potential to help patients with AD objectively assess their skin condition, facilitating timely and appropriate treatment." They added, "This study lays the groundwork for future advancements in AI-driven dermatological assessments, enhancing both patient care and clinical research." The development of such tools promises a future where managing skin conditions like atopic dermatitis becomes more precise, accessible, and patient-centered.

References:

1. Okata-Karigane U, Hirota M, Takahashi C, et al. AI-based objective severity assessment of atopic dermatitis using patient photos in a real-world setting: A digital biomarker approach. Allergy. Published online May 19, 2025. doi:10.1111/all.16586
2. AI Tool Enables Real-World Assessment of Eczema Severity via Smartphone Photos. News release. Keio University School of Medicine. May 20, 2025. Accessed May 30, 2025. https://www.keio.ac.jp/en/press-releases/2025/May/20/49-167063/?utm_source=imaginepro.ai

The original article mentioned a link to atopic dermatitis for further information.