Development of a digital image analysis system for objective evaluation of treatment response in cellulitis
Authors:
Open Access
- 17 December 2025
- journal article
- Published by Springer Nature in Scientific Reports
- Vol. 15 (1) , 44032
Abstract
Cellulitis is a common skin infection. While changes in lesion color are valuable indicators of treatment efficacy, assessments often rely on subjective judgments. To provide an objective method for evaluating treatment response, we developed a digital imaging system that quantifies red, green, and blue (RGB) values of lesions and calculates the RGB distance between lesions and normal skin. We analyzed 34 patients with cellulitis treated between 2022 and 2025. Clinical images were obtained on days 1 (baseline), 3, 7, and 14. By day 14, the G and B values of lesions increased significantly (G: 115.6 to 135.7, B: 91.8 to 105.5; P < 0.05), while the R values remained stable. The RGB distance decreased significantly from day 3 (64.3 to 53.7; P < 0.05). Patients were classified by Investigator's Global Assessment into Group A (favorable clinical outcome, n = 13) and Group B (delayed clinical improvement, n = 21). Receiver operating characteristic analyses of the RGB distance ratio relative to day 1 showed high accuracy at day 3 (cut-off = 0.62) and day 7 (cut-off = 0.48). A significantly higher proportion of Group B patients had a history of edema. These findings suggest our system provides an objective method for evaluating the clinical course of cellulitis.Keywords
Funding Information
- The Ministry of Education, Culture, Sports, Science and Technology of Japan (22K08416)
- Health and Labor Sciences Research Grants for Research on Measures for Intractable Diseases from Ministry of Health, Labor and Welfare of Japan (23FC1039)
This publication has 29 references indexed in Scilit:
- Cellulitis: A Review of Current Practice Guidelines and Differentiation from PseudocellulitisAmerican Journal of Clinical Dermatology, 2021
- Classification of large-scale image database of various skin diseases using deep learningInternational Journal of Computer Assisted Radiology and Surgery, 2021
- CellulitisMedical Clinics of North America, 2021
- Artificial intelligence-based image classification methods for diagnosis of skin cancer: Challenges and opportunitiesComputers in Biology and Medicine, 2020
- Compression Therapy to Prevent Recurrent Cellulitis of the LegNew England Journal of Medicine, 2020
- Costs and Consequences Associated With Misdiagnosed Lower Extremity CellulitisJAMA Dermatology, 2017
- CellulitisJAMA, 2016
- Clinical and epidemiological characteristics of adult patients hospitalized for erysipelas and cellulitisEuropean Journal of Clinical Microbiology & Infectious Diseases, 2012
- Cellulitis: Definition, Etiology, and Clinical FeaturesThe American Journal of Medicine, 2011
- Skin Melanin, Hemoglobin, and Light Scattering Properties can be Quantitatively Assessed In Vivo Using Diffuse Reflectance SpectroscopyJournal of Investigative Dermatology, 2001