Comparison of Dermatoscopy X400 Interpreted by 6 Non-dermatologists and an Artificial Intelligence Software Skin Artificial Intelligence Versus Dermatoscopy x 20

Completed

• Conditions: Pigmented Lesions; Genital Tumor
• Interventions: Other: Analysis of photos of pigmented tumours by the three groups

• Registration Number: NCT06079619
• Lead Sponsor: Centre Hospitalier Universitaire de Saint Etienne

Brief Summary

Pigmented lesions of the genital mucosa are common, affecting between 10% and 20% of the population. The clinical appearance is sometimes confusing. They are often extensive, irregular, and inhomogeneous in color, which can make them look suspicious. Moreover, dermatoscopic imaging is often complex to interpret, as shown by the many different patterns reported in the literature.

Thus, preliminary work in confocal microscopy has shown that it is possible to simplify diagnosis by using microscopic, in vivo, non-invasive imaging. However, CIM cameras are rare, very expensive and difficult to access.

Conversely, digital dermatoscopes with 400x magnification are much less expensive and can also allow, in a certain number of cases, an examination at the cellular level of tumors, with, in our opinion, the advantage of being easy to learn. In the context of the pooling of resources within medical centers in particular, their dissemination should accelerate. Furthermore, x400 dermatoscopy can provide similar information to IVCM for the most frequent black genital tumors.

Detailed Description

The hypothesis that motivates this work is that X400 dermatoscopy in the hands of non-experts would allow a fast increase in their diagnostic capacity by identifying the main genital pigmented tumor, namely melanosis, and thus improve the use of expert centers relieved of the care of benign tumors.

Study Timeline

• Study Start: 2022-03-01
• Primary Completion: 2022-04-20
• Study Completion: 2022-04-20
• Status Verified: 2023-09
• Study First Submitted: 2023-10-06
• Study First Submitted QC: 2023-10-06
• Study First Posted: 2023-10-12
• Last Update Submitted: 2024-04-03
• Last Update Posted: 2024-04-05

Recruitment & Eligibility

• Status: COMPLETED
• Sex: All
• Target Recruitment: 157

Inclusion Criteria

• Men and women with a pigmented lesion of the genital mucosa, having benefited from a mucosal consultation at the University Hospital of Saint-Etienne, with Professor Perrot.
• The patient's clinical file must contain clinical images, dermatoscopy images x 20 and at least 1 dermatoscopy image x 400.

Exclusion Criteria

• None

Study & Design

• Study Type: OBSERVATIONAL
• Study Design: Not specified

Arm && Interventions

Group	Intervention	Description
Group C: Artificial intelligence	Analysis of photos of pigmented tumours by the three groups	Artificial intelligence: the skin Artificial intelligence software, a learning base will be provided, different from the evaluation base and the test base.
Group B: non-dermatologists	Analysis of photos of pigmented tumours by the three groups	Group B is composed of gynecologists (Group B1) and general practitioners (B2) Group B physicians will only perform dermatoscopy x 400. They will have been given a maximum of 1 hour's training to identify melanosis-type pigmented tumors, which are the most frequent benign tumors in x 400 dermatoscopy. Only three x400 dermatoscopy patterns will be presented to them to the exclusion of all others, which are sufficient to identify benign lesions.
Group A: expert dermatologists	Analysis of photos of pigmented tumours by the three groups	Group A is composed of expert dermatologists. Group A doctors will have at their disposal 150 dermoscopic photos, x20 dermatoscopic photos according to the usual practice of pigmented genital tumors. The dermatologists will characterize the dermatoscopy with a standardized reading grid to assess the different dermatoscopic patterns of genital tumors.

Primary Outcome Measures

Name	Time	Method
Comparison results between expert dermatologists with x20 dermatoscopic photos (Group A) and non-dermatologists physicians with x400 dermatoscopic photos (Group B)	Year: 1	Performance of x400 dermatoscopic (results of expert dermatologists versus non-dermatologists physicians)

Secondary Outcome Measures

Name	Time	Method
Comparison results between expert dermatologists with x20 dermatoscopic photos (Group A) and gynecologists with x400 dermatoscopic photos (Group B1)	Year: 1	Performance of x400 dermatoscopic (results of expert dermatologists versus gynecologists)
Comparison results between gynecologists with x400 dermatoscopic photos (Group B1) and general practitioners with x400 dermatoscopic photos (Group B2).	Year: 1	Performance of x400 dermatoscopic (results of gynecologists versus general practitioners)
Comparison results between expert dermatologists with x400 dermatoscopic photos (Group A) and skin Artificial intelligence software with x400 dermatoscopic photos (Group C).	Year: 1	Performance of x400 dermatoscopic (results of expert dermatologists versus skin Artificial intelligence software)
Comparison results between expert dermatologists with x20 dermatoscopic photos (Group A) and general practitioners with x400 dermatoscopic photos (Group B2)	Year: 1	Performance of x400 dermatoscopic (results of expert dermatologists versus general practitioners)
Comparison results between dermatologists physicians with x400 dermatoscopic photos (Group B) and skin Artificial intelligence software with x400 dermatoscopic photos (Group C).	Year: 1	Performance of x400 dermatoscopic (results of dermatologists physicians versus skin Artificial intelligence software)

Trial Locations

Locations (1)

CHU Saint Etienne; 🇫🇷 Saint-Étienne, France