AI-Driven Tool for Melanoma Diagnosis Displays Promising Results
Dermalyser shows the potential to decrease the workload for dermatologists and replace it with an artificial intelligence tool to identify malignant melanomas among cutaneous lesions.
A positive diagnostic performance was seen with the Dermaslayer device which is used as a decision support system for assessing potential melanoma. Findings included 95% sensitivity and 86% specificity in the first prospective study of its kind conducted in primary care settings, according to AI Medical Technology (NCT05172232).1
“Indeed, [these are] exciting results, these numbers show potential of not only improving future visual diagnostic accuracy, but also decreasing the amount of workload that dermatologists too often are dealing with in their daily practice. Nevertheless, additional studies are necessary to confirm the positive results. Human-computer collaboration is here to stay and I'm looking forward to its implementation in the field of skin cancer and dermatology in general,” added Olle Larkö, professor in Dermatology & Venereology and former Dean at Sahlgrenska University, in the press release.
The clinical trial took place across 37 Swedish primary care facilities and evaluated a diagnostic decision support system empowered with advanced artificial intelligence called Dermalyser. Findings showed that Dermalyser outperformed primary care physicians and dermatologists.
In the trial, a total of 240 patients were enrolled who were seeking primary care for their cutaneous lesions that they suspected to be melanoma. The prospective, pivotal, confirmatory, multi-center, non-interventional trial assessed the primary end point of diagnostic accuracy to detect melanoma and secondary end point of instrument usability of the device.
This is the first study of its kind in this setting to examine the clinical safety, performance, and benefit of Dermalyzer in patients with cutaneous lesions where malignant melanoma cannot be ruled out.
The trial included patients who are being seen by primary care physicians who have melanoma suspicious skin lesion(s). Patients are treated with the diagnostic test, Dermalyzer, to evaluate the diagnostic accuracy of the device to detect melanoma among cutaneous skin lesions by dermoscopy.
Patients included in the trial were those aged 18 years and older who were attending a primary care facility and had at least 1 suspicious skin lesion where malignant melanoma could not be ruled out.2
Those with cutaneous lesions considered as benign, those in areas that are not suitable for dermoscopy imaging, and those in areas with any form of scarring of tissue due to injury are excluded from the study. Other exclusion criteria included having damaged or injured non intact skin where the cutaneous lesion is located, type V and VI skin defined by the Fitzpatrick scale, cutaneous lesions in areas that are covered by tattoos, and those in areas covered by abundantly hairy skin. Additionally, those with images where the entire lesion is not inside the photo or images that are out of focus were not included from the study.
"The remarkably high sensitivity and specificity levels demonstrate the clinical performance and benefit of Dermalyser, particularly since the study was conducted in a real world, primary care setting representing different demographics, personnel, and geographical location", said Christoffer Ekström, chief executive officer of AI Medical Technology, in the press release. "We now look forward to finalizing the CE marking and preparing for market launch in Europe [by the end of] 2023. Simultaneously, we plan to engage [United States]-based clinics in upcoming clinical trials as part of our [United States] market entry strategy".
"In addition to the primary objective, i.e., the diagnostic precision of the device, we also determined the usability and applicability of Dermalyser in clinical practice. Results from the study will be reported in a clinical publication, but already the clear impression is a high acceptance of Dermalyser as a diagnostic support tool among the medical personnel involved,” added Magnus Falk, associate professor in general practice and principal investigator, Department of Health, Medicine and Caring Sciences, Linköping University, in the press release.
