A new artificial intelligence-powered smartphone application has been developed to help detect deadly skin cancers, potentially reducing National Health Service (NHS) waiting lists. The software, named Derm Zero, analyzes photographs of moles and skin lesions to identify patterns associated with skin cancer and other conditions.
Developed by the British healthcare firm Skin Analytics, the application has recently received the highest level of medical device approval in Europe. This allows patients to receive clinical-grade skin checks within seconds at pharmacies or general practitioner offices, without requiring a hospital appointment. An earlier version of the technology, known as Derm AI, was used by the health service to detect 20,000 cancers in over 230,000 patients. However, that version required a special camera lens attachment, which the latest iteration does not need.
The Derm Zero software has been trained on thousands of images with confirmed diagnoses. It uses AI to distinguish between benign lesions and those that require clinical review. According to the developers, the software has an accuracy rate of 98.9 percent, meaning it correctly identifies skin cancer in approximately 97 out of every 100 patients.
Melanoma, the most aggressive form of skin cancer, affects around 20,000 people in the UK annually and is responsible for more than 2,300 deaths each year. Rates of melanoma have increased significantly in the past decade, reaching a record high last year. The primary indicator of melanoma is the appearance of a new mole or changes in the size or shape of an existing one. While melanomas can develop anywhere on the body, they are most common in areas frequently exposed to the sun.
Dr. Alexandra Kemp, consultant dermatologist and cancer clinical director at Amersham Hospital, stated that the technology has already had a significant impact on clinical capacity and care efficiency. "When it comes to skin cancer, we know early diagnosis means better outcomes," Kemp said. "To see this technology available on a smartphone, with no specialist equipment required, creates huge possibilities for expanding patient access and enabling earlier diagnosis."