AI Three-Item Severity Score Can ID Eczema From Patient-Uploaded Images

FRIDAY, May 23, 2025 (HealthDay News) — An artificial intelligence (AI) Three-Item Severity (TIS) score can identify eczema-affected areas and severity scores from user-uploaded images, according to a research letter published online May 19 in Allergy.

Utako Okata-Karigane, M.D., from the Keio University School of Medicine in Japan, and colleagues developed an AI-based model to assess eczema severity from user-uploaded photos. Three algorithms were trained and integrated: body part detection, skin lesion detection, and severity assessment. The TIS score, a localized severity assessment ranging from 0 to 9, evaluating erythema, edema/papulation, and excoriation, was adopted because the model analyzes representative images rather than full-body images.

A total of 9,656 images with itch scores were included to establish and validate AI-TIS. The researchers found that the trained AI model correctly detected 98 and 100 percent of body parts and eczema areas, respectively; across 220 test images, the model correlated well with TIS scores determined by board-certified dermatologists and/or allergists (R = 0.73). Across 8,556 images, AI-TIS showed a lower correlation with itch intensity on a 0-to-5 numeric rating scale (R = 0.11). In 15 patients, correlations of AI-TIS with TIS, SCORing Atopic Dermatitis (SCORAD), and objective-SCORAD were R = 0.16, 0.4, and 0.53, respectively.

“This study lays the groundwork for future advancements in AI-driven dermatological assessments, enhancing both patient care and clinical research,” the authors write.

Several authors disclosed ties to the biopharmaceutical industry.

Abstract/Full Text