Clinical evidence
This section compiles the peer-reviewed publications behind the automated atopic dermatitis scoring provided by Legit.Health, together with the foundational SCORAD and EASI references. These papers document the scientific basis and real-world use of the automated SCORAD and EASI algorithms.
Validation status and reproducibility
The automated SCORAD engine was validated in a peer-reviewed pilot study (Medela et al., JID Innovations, 2022), which showed that automatic scoring of the intensity signs reached agreement comparable to expert dermatologist assessment while reducing the inter-observer variability that limits manual SCORAD.
Severity scoring has no objective gold standard: the reference is the mathematical consensus of independent expert dermatologists, the same standard the FDA and EMA accept for dermatology trials. In this study the ground truth was built from three expert dermatologists per dataset (nine across the study's three datasets). Matching that consensus is the realistic performance ceiling for any rater, human or AI.
Two properties are decisive for endpoint use in a trial:
- Reproducibility: scoring is deterministic. The same image yields the same score at every site and every visit, with no calibration drift and no inter-reader variability, the dominant source of noise in multi-centre severity endpoints.
- Shared, validated architecture: the same segmentation and intensity-scoring engine underpins the platform's other automated indices (Automated PASI for psoriasis, automated SALT for alopecia, and automated acne scoring), which have been evaluated on larger datasets, reinforcing the overall evidence base.
The device is CE-marked as a medical device, meaning a Notified Body has independently assessed it against the safety and performance requirements of the EU Medical Device Directive (MDD 93/42/EEC) and authorised its use on the EU market. Beyond the EU, the device is registered with the MHRA for the United Kingdom market and has obtained ANVISA approval in Brazil. This is a higher bar than general-purpose software: the intended use, the clinical evidence, and the ongoing performance monitoring are all regulated. Its automated scores are used as endpoints in clinical research. Real-world performance is monitored continuously through the manufacturer's post-market surveillance and post-market clinical follow-up (PMCF) programme. Study-specific agreement and reproducibility metrics are available on request to support protocol design.
Legit.Health is a clinical decision support device: the automated scores provide diagnostic support and do not replace the healthcare professional's assessment.
Automated SCORAD
"Automatic SCOring of Atopic Dermatitis using deep learning: A pilot study" Medela, A., Mac Carthy, T., Aguilar Robles, S. A., Chiesa-Estomba, C. M., Grimalt, R. JID Innovations, Volume 2, Issue 3, 100107 (2022) DOI: 10.1016/j.xjidi.2022.100107 | PMID: 35990535
A collaboration with Dr. Ramon Grimalt, published in JID Innovations, describing the deep learning approach that automates SCORAD severity assessment of atopic dermatitis from photographs.
SCORAD takes 7–10 minutes to score by hand, and because it is complex, well-trained dermatologists can grade the same case differently. Automated SCORAD returns an objective severity read from photographs that is comparable to expert assessment while reducing inter-observer variability, a major source of noise in multi-centre atopic dermatitis endpoints.
This pilot established proof of concept for automated SCORAD, and the technology has advanced considerably since. Two changes matter most for a trial today.
First, the affected body surface area (BSA) is now measured automatically from the images through pixel-level segmentation, instead of the manual rule-of-nines estimate the original pilot still relied on. Because visual BSA estimation is the single largest source of inter-rater variability in manual SCORAD and EASI, measuring it objectively removes the most operator-dependent step of the score and tightens the endpoint across sites and visits.
Second, the training dataset has been substantially expanded and diversified across skin tones. Erythema and other inflammatory signs present differently on darker skin, and underrepresented phototypes are a well-documented blind spot for dermatology AI, so this is a deliberate, evidence-backed extension rather than an incidental one. As a result, the platform now detects and grades atopic dermatitis on darker skin tones (Fitzpatrick V–VI) with performance comparable to expert dermatologist assessment. See Performance across skin types for metrics stratified by Fitzpatrick group.
Medela, A., Mac Carthy, T., Aguilar Robles, S. A., Chiesa-Estomba, C. M., & Grimalt, R. (2022). Automatic SCOring of Atopic Dermatitis Using Deep Learning: A Pilot Study. In JID Innovations (Vol. 2, Issue 3, p. 100107). Elsevier BV. https://doi.org/10.1016/j.xjidi.2022.100107
Image quality for atopic dermatitis scoring (DIQA)
"Dermatology Image Quality Assessment (DIQA): Artificial intelligence to ensure the clinical utility of images for remote consultations and clinical trials" Hernández Montilla, I., Mac Carthy, T., Aguilar, A., Medela, A. Journal of the American Academy of Dermatology, Volume 88, Issue 4, pp. 927–928 (2023) DOI: 10.1016/j.jaad.2022.11.002 | PMID: 36526082
Reliable SCORAD and EASI scoring depends on the quality of the input photographs. DIQA is the image quality assessment algorithm that acts as a quality gate in the atopic dermatitis imaging workflow, checking every image against consistent quality criteria across investigator sites before it reaches the scoring algorithms.
Image quality is the hidden failure point of decentralised and multi-site imaging. DIQA screens every photograph for clinical utility before it reaches the scoring algorithms, so unusable images are caught at capture rather than surfacing as missing data at database lock.
Hernández-Montilla, I., Mac Carthy, T., Aguilar, A., & Medela, A. (2023). Dermatology Image Quality Assessment (DIQA): Artificial intelligence to ensure the clinical utility of images for remote consultations and clinical trials. In Journal of the American Academy of Dermatology (Vol. 88, Issue 4, pp. 927–928). Elsevier BV. https://doi.org/10.1016/j.jaad.2022.11.002
Foundational SCORAD and EASI references
The platform does not introduce a new, unvalidated scale. It automates SCORAD and EASI, the severity indices clinicians and regulators already accept, so the endpoint your protocol specifies is the endpoint the algorithm reports.
The scoring systems automated by the platform are the established clinical indices for atopic dermatitis:
- Severity scoring of atopic dermatitis: the SCORAD index. Dermatology, 186(1), 23–31 (1993). The original SCORAD development paper.
- Oranje, A. P. et al. Practical issues on interpretation of scoring atopic dermatitis: the SCORAD index, objective SCORAD and the three-item severity score. British Journal of Dermatology, 157(4), 645–648 (2007).
- Hanifin, J. M. et al. The eczema area and severity index (EASI): assessment of reliability in atopic dermatitis. Experimental Dermatology, 10(1), 11–18 (2001).
For the full list of clinical evidence across all indications, see the clinical validation section.