AI Medical Technology announced that Dermalyser, its AI-powered smartphone application for melanoma detection, has received CE mark approval for use across the European Union.
The company noted that it plans to expand its footprint across Europe over the coming year, with a US market launch scheduled for 2027. Adding, the decision clears the way for the wide adoption of the tool that clinical studies show significantly outperforms even experienced dermatologists in identifying melanoma, the deadliest form of skin cancer.
The World Health Organization (WHO) identifies skin cancers as the most common group of cancers diagnosed worldwide. Early detection is critical, as most of the melanoma cases are curable when diagnosed early.
Dermalyser, classified as a stand-alone Medical Device Software (MDSW), uses a dermatoscope attached to a smartphone to capture high-resolution images of skin lesions. These are then analyzed in real time by an AI algorithm trained to detect melanoma with high accuracy.
Real world data
In June 2024, AI Medical Technology published the results of a prospective multicentre clinical trial in the British Journal of Dermatology. The researchers evaluated Dermalyser’s performance in a real-world clinical setting across 36 primary care centers in Sweden. The prospective trial included 253 suspicious lesions from 228 patients. Of these, 21 were confirmed as melanomas—11 thin invasive melanomas and 10 melanomas in situ.
Dermalyser correctly identified 20 of the 21 melanomas, missing only one in situ case. The app achieved an area under the receiver operating characteristic (AUROC) curve of 0.960 for all melanomas and 0.988 for invasive melanomas. This corresponds to near-perfect sensitivity and very high specificity, especially when compared to expert dermatologists (AUROC of approximately 0.85) and general practitioners (AUROC of about 0.70).
Dermalyser achieved a 95.2% sensitivity and 84.5% specificity rate in clinical trials. For invasive melanomas, the most dangerous subtype, the tool demonstrated 100% sensitivity and 92.6% specificity.
The app also demonstrated a 99.5% negative predictive value (NPV), indicating its effectiveness in ruling out melanoma and reducing unnecessary surgical excisions of benign lesions. When combined with a physician’s clinical judgment, Dermalyser improved the overall diagnostic accuracy further, underscoring its potential as a decision-support tool in primary care.
Study authors concluded that integrating AI tools like Dermalyser into everyday clinical practice can significantly enhance early melanoma detection and optimize referral decisions—potentially transforming how skin cancer is diagnosed and managed at the front line of care.
