Developments in AI for detecting skin cancers

detecting skin cancers

Artificial Intelligence (AI) is making waves in the medical field, particularly in detecting skin cancers. A recent study into an AI-driven tool, known as the Deep Ensemble for the Recognition of Malignancy (DERM), has proven highly accurate in identifying two common forms of skin cancer – squamous cell carcinoma (SCC) and basal cell carcinoma (BCC).

This smart tool isn’t just good at spotting cancerous lesions; it can also help manage non-cancerous ones too. So, could AI be the key to detecting and treating skin cancer early? Here, we look at how AI can aid a dermatologist when detecting skin cancers.

AI-power for early detection of skin cancers

UK doctors carried out a study to determine whether the DERM tool could be successful in detecting skin cancers. A total of 572 participants were included, who were primarily elderly patients, across four NHS centres. Anyone with a suspicious skin mark that could be photographed was included. These marks were captured using smartphones that were each equipped with a special lens for skin examination.

The study collected images of 592 skin marks, and for 395 of these marks, a skin tissue test result was available for comparison. The marks were mostly found on the head and scalp, legs, arms, and back. A total of 176 basal cell carcinoma, 44 squamous cell carcinoma, and 297 non-cancerous marks were detected.

The doctors also categorised the chance of each mark being skin cancer, from ‘unlikely’ to ‘highly likely’. The AI tool analysed these images, and its performance was consistent across all three smartphones, despite each one capturing slightly different pictures due to factors like lighting variations.

Could AI be the key to addressing dermatology shortages?

AI technology could be a game-changer for dermatologists currently struggling with their workflow. Referrals for potential skin cancers have doubled over the past decade to over 700 per 100,000 people. However, the number of dermatologists hasn’t risen at the same level. This has led to a shortage in the profession, with current dermatologists struggling to keep up with demand.

The situation is exacerbated by an ageing population, many of whom are sun-lovers. This is where AI steps in. By accurately identifying both cancerous and non-cancerous skin lesions, AI could help dermatologists prioritise cases and streamline their workflow. This could, in turn, enable faster diagnosis and treatment for patients, an essential factor in effective skin cancer management.

Do I still need to see a Dermatologist?

While AI technology for skin cancer detection is making headway, it cannot yet match the skill and experience of an experienced dermatologist.

Dermatologists are equipped with years of medical training and an intricate understanding of skin health. They can quickly detect early skin cancer symptoms, some of which are not yet recognisable by AI. They can also quickly organise treatment if skin cancer is detected, increasing the likelihood of a successful outcome. From biopsies to removing cancerous lesions, their immediate interventions can be lifesaving.

If you notice any unusual changes in your skin, or you have any skin concerns, schedule an appointment with Consultant Dermatologist Dr. Juliet Williams straight away by calling 01483 555907.