An artificial intelligence (AI) tool that checks suspicious skin spots has been conditionally approved for use in NHS services in England. The approval comes from the National Institute for Health and Care Excellence (NICE).
The tool is called Deep Ensemble for Recognition of Malignancy (DERM). It can now be used while more research is done to confirm its benefits.
The goal of using DERM is to help patients get quicker checks for possible skin cancer. Right now, about one in three patients are not seen within the NHS’s 18-week target for urgent skin cancer referrals.
How DERM Works
DERM helps with teledermatology, which is a way to check skin problems remotely after a patient is referred by a GP.
Here’s how it works:
- Healthcare staff take a close-up photo of the skin spot using a smartphone with a special lens.
- The photo is uploaded to DERM’s secure system.
- DERM uses AI to compare the image to a set database of known skin conditions. It does not learn from new data but works from a fixed set of examples.
It checks for:
- Melanoma
- Squamous cell carcinoma
- Basal cell carcinoma
- Intra-epidermal carcinoma
- Actinic keratosis
- Atypical moles (nevus)
- Non-cancerous spots like seborrheic keratosis and sun spots
DERM then gives a possible diagnosis and helps decide how urgent the case is:
- Non-serious cases may get written advice and not need urgent care.
- Serious or unclear cases are sent to a dermatologist for review online.
NICE said DERM may be as accurate as seeing a skin specialist in person or through video, and it can help reduce the workload in dermatology departments.
Source: medscape
