How Consistent is the Output of DERM, When Used to Assess Images of Potentially Cancerous Skin Lesions.

NCT06654999 | Completed

• 1 other identifier: DERM-011
• Study Type: observational
• Target: 115
• Locations: 1 country
• Sites: 3

Brief Summary

DERM is a Medical Device that uses artificial intelligence to help doctors check if a skin lesion might be cancerous. It works by analysing close-up pictures of skin lesions taken with a smartphone. This study aims to demonstrate how consistent (precise) the output of DERM is: i.e. does it provide the same result when it analyses multiple photos of the same lesion (repeatability), and when the same lesion is photographed by different people, or with different cameras (reproducibility). Adults with at least one skin lesion that doctors are checking for cancer, as part of their standard care, will be able to take part. Suitable lesions will be photographed three times, each by three different people using three sets of image capture hardware (specifically, an iPhone 11 with a DL200/HR dermoscopic lens). Each image will be checked for good image quality as it is captured. Images will then be transferred to DERM, where they'll be analysed. The DERM output won't be shared with the patients or doctors involved in the study. The patients will continue to have their skin lesion biopsy/excised, in accordance with standard of care. Their diagnosis will be collected and compared to the output from DERM.

Conditions

Skin Cancer; Melanoma of Skin; Squamous Cell Cancer; Basal Cell Cancer

Keywords

Precision; DERM; Repeatability; Reproducability

Outcome Measures

Primary Outcomes (2)

• Reproducibility: Average Positive Agreement (APA) of DERM on malignant lesions is >80%

  Each lesion will be imaged by three users using three sets of the image capture hardware, generating nine data sets (user/hardware combinations) per lesion. Each lesion image data set will consist of 3 repeated measures, resulting in 27 measures per lesion

  1 day

• Repeatability: Average Positive Agreement (APA) of DERM on malignant lesions is >80%

  Each lesion will be imaged by three users using three sets of the image capture hardware, generating nine data sets (user/hardware combinations) per lesion. Each lesion image data set will consist of 3 repeated measures, resulting in 27 measures per lesion.

  1 day

Secondary Outcomes (2)

• Reproducibility: Average Negative Agreement (ANA) of DERM on benign lesions is >50%

  1 day

• Repeatability: Average Negative Agreement (ANA) of DERM on benign lesions is >50%

  1 day

Study Arms (1)

All patients

All patients will have their skin lesion imaged in the same way

Device: Deep Ensemble for the Recognition of Malignancy (DERM)

Interventions

Deep Ensemble for the Recognition of Malignancy (DERM) DEVICE

DERM variants "+" and "DS"

All patients

Eligibility Criteria

• Age: 18 Years+
• Sex: all
• Healthy Volunteers: No
• Age Groups: Adult (18-64), Older Adult (65+)
• Sampling Method: Non-Probability Sample

Study Population

Patients will be recruited in specialist dermatology clinics in the UK.

You may qualify if:

• Male or Female,
• aged 18 years or above,
• Willing and able to give informed consent for participation in the study,
• Has at least one skin lesion that, following assessment by a primary care provider and a dermatology specialist, will be biopsied due to a high suspicion of skin cancer, and is suitable for assessment by DERM.
• In the Investigator's opinion, able and willing to comply with all study requirements.

You may not qualify if:

• Patients with more than one skin lesion due for biopsy
• Any other significant disease or disorder which, in the opinion of the Investigator, may either put the participant at risk because of participation in the study, or may influence the result of the study, or the participant's ability to participate in the study.
• Be located on an anatomical site of different skin structure: palms of hands or soles of feet (acral lesion), mucosal surfaces (lips and eyes) or under nail (ungal lesion),
• Have a diameter greater than the diameter of the dermoscopic lens,
• Be located on an anatomical site unsuitable for photographing, including on surface of genitals and hair-bearing areas,
• Have been previously biopsied, excised, treated or is otherwise traumatised
• Be located in an area of visible scarring or tattooing.

Sponsors & Collaborators

• Skin Analytics Limited: lead

Study Sites (3)

Norwich and Norfolk Hospitals Trust

Norwich, United Kingdom

Location

University Hospitals Dorset

Poole, United Kingdom

Location

Surrey and Sussex Hospital Trust

Redhill, United Kingdom

Location

MeSH Terms

Conditions

Skin Neoplasms; Melanoma; Neoplasms, Squamous Cell; Neoplasms, Basal Cell

Condition Hierarchy (Ancestors)

Neoplasms by Site; Neoplasms; Skin Diseases; Skin and Connective Tissue Diseases; Neuroendocrine Tumors; Neuroectodermal Tumors; Neoplasms, Germ Cell and Embryonal; Neoplasms by Histologic Type; Neoplasms, Nerve Tissue; Nevi and Melanomas; Neoplasms, Glandular and Epithelial

Study Design

• Study Type: observational
• Observational Model: CASE ONLY
• Time Perspective: PROSPECTIVE
• Sponsor Type: INDUSTRY
• Responsible Party: SPONSOR

Study Record Dates

• First Submitted: October 22, 2024
• First Posted: October 23, 2024
• Study Start: December 18, 2024
• Primary Completion: June 30, 2025
• Study Completion: June 30, 2025
• Last Updated: July 9, 2025

Record last verified: 2025-07

Locations

GB (3)