Saltar al contenido principal

22 documentos etiquetados con "Clinical evidence"

Ver Todas las Etiquetas

Camera calibration

How to calibrate a camera model before study imaging, including the calibration board, camera requirements, capture procedure, and submission process.

Clinical evidence

Evidence behind automated SALT scoring for alopecia, covering the validation of the hair loss segmentation model, the DIQA image quality gate, conference presentations, and the foundational SALT references.

Clinical evidence

Peer-reviewed publications behind automated SCORAD and EASI scoring for atopic dermatitis, including the primary automated-SCORAD study and the foundational SCORAD and EASI references.

Clinical evidence

Peer-reviewed publication behind automated IHS4 scoring for hidradenitis suppurativa, covering the AIHS4 study, the DIQA image quality gate, and the foundational IHS4 and HiSCR references.

Clinical evidence

Peer-reviewed publications and production validation behind automated PASI scoring for psoriasis, covering the peer-reviewed validation study, the DIQA image quality gate, and per-component performance against expert inter-rater variability.

Clinical Evidence and Validation

Performance validation of acne severity endpoints, covering how the ground truth is established, what the metrics mean, ALADIN vs. dermatologist agreement, lesion detection performance, and the regulatory evidence pathway.

Clinical Trial Workflow and Data Integration

End-to-end clinical trial workflow for atopic dermatitis severity endpoints, covering protocol design, patient enrolment, image capture, AI scoring, severity tracking, data export, and EDC integration.

Cutaneous Lupus Erythematosus Severity Endpoints for Clinical Trials

Automated CLASI scoring for cutaneous lupus erythematosus clinical trials. AI quantification of the visual components of CLASI Activity and CLASI Damage across 13 anatomical regions, plus a continuous repigmentation measure, from smartphone images. CE-marked medical device.

Image capture

Standardised steps for site staff to place calibration markers and capture lesion photographs, with a do and do-not checklist, how the number of detected markers is reported back, the printable one-page capture guide issued per project, correct and incorrect placement examples by condition, troubleshooting, and storage guidance.

index

How calibration markers give the Legit.Health platform a known reference for measuring lesion surface area, and for calibrating colour across visits in dermatology clinical trials.

Known Limitations

Transparent documentation of what the AI alopecia severity scoring can and cannot do, including segmentation challenges, validation status, and how each limitation is managed.

Known Limitations

Transparent documentation of what the automated SCORAD and EASI scoring can and cannot do, including the inherent constraints of photograph-based assessment and how each limitation is managed.

Marker specifications

The physical specification of the Legit.Health Color Multi-Marker, its black and white borders, the 5 by 5 colour-patch grid, and the requirements a marker must meet for reliable automated detection.

Publications, References, and Further Reading

Peer-reviewed publications, conference presentations, key references from the dermatology literature, and the complete clinical validation portfolio supporting acne severity endpoints.

Scoring Methodology: How the AI Calculates PASI

Technical detail on the automated PASI pipeline, covering affected-area segmentation, intensity scoring of erythema, desquamation and induration, the PASI formula, the severity scale, and comparison with manual scoring.

Scoring Methodology: SALT for Alopecia

Technical detail on how pixel-level hair loss segmentation produces the SALT score for alopecia, covering the four scalp quadrants and their weights, the SALT formula, severity bands, responder definitions against baseline, a worked example, and comparison with manual assessment.

Worked Example: From Capture to Scored Report

The psoriasis assessment sequence step by step, from the eleven capture screens the subject sees, through the images they submit, to the segmentation overlays the device returns, the scored PASI report format, and the automatic response-milestone alert.