Related Experiment Video
Updated: Sep 9, 2026
SCAnED - An Open-source Skin Segmentation Macro for Semi-automated Cell and Nuclei Detection in Epidermal and Dermal Skin Compartments
Published on: August 8, 2025
Dermoscopic CIELab Colorimetry for Longitudinal Monitoring of Fixed Target Lesions in Psoriasis: Objective
Ziyuan Tian1,2, Yang Han3, Zhou Zhuang1,4
1Department of Dermatology, China-Japan Friendship Hospital, Beijing, People's Republic of China.
Purpose:
Objective longitudinal monitoring of a single psoriasis lesion remains underdeveloped. We evaluated the methodological feasibility of a quantitative fixed-target-lesion framework that derives continuous surface-color and vascular signals from dermoscopy, using CIELab as the measurement space rather than the clinical endpoint.
Patients And Methods:
This single-center retrospective longitudinal study included 28 patients with moderate-to-severe plaque psoriasis receiving biologic therapy (176 dermoscopic observations). Linear mixed-effects models (LMM) tested the longitudinal association between a* and PASI, and cumulative logit mixed models (CLMM) tested associations with ordinal dermoscopic scores. Data-driven a* thresholds were then applied unchanged to a separate mild-psoriasis cohort from the same center, device platform, and institutional database (BW cohort; 14 patients, 42 visit-level observations with concurrent scores) as a within-center transferability assessment.
Results:
a* was positively associated with PASI in the random-slope LMM (β = 0.928, p < 0.001). Data-driven a* tiers showed moderate agreement with expert background-color grading in the primary cohort (quadratic weighted κ = 0.538) and higher agreement in the unchanged-threshold BW assessment (κ = 0.722; ±1-level agreement 95.2%). The BW analysis was interpreted as within-center transferability, not external validation. Automated vessel-signal counts were associated with ordinal vessel-density scores (CLMM OR = 2.53 per SD, p < 0.001) and remained associated after adjustment for a* (OR = 2.13 per SD, p < 0.001). Exploratory treatment-group findings were treated as hypothesis-generating only.
Conclusion:
These findings provide proof-of-concept support for an objective, quantitative framework for monitoring fixed psoriatic target lesions. a* and rule-based vessel-signal counts are candidate lesion-level measures that may complement conventional scores, but clinical utility and cross-center or cross-device generalizability remain unestablished.
