Artificial intelligence technology can help to meet rising demand for early detection of melanoma.
Skin cancer and melanoma (the most severe type of skin cancer) are becoming a social health issue. The incidence has been rising. Currently, between 2 and 3 million non-melanoma skin cancers and 132,000 melanoma skin cancers occur globally each year.
Experts agree that early detection is essential and can save lives. One third of those diagnosed are below 50 at the time of diagnosis. There is still no medical cure, but recently some new treatments have emerged that can help to extend life for few years.
Those new treatments for advanced melanoma come at a cost, but, if detected at an early stage and removed by excision, the cost is more than 100 times less.
Early detection starts with population awareness – people should always consult their doctor if they are in any doubt about a mole. However, the trouble is that early-stage melanoma can easily be confused with benign moles, and 90% of the population has at least one mole.
With cases on the rise, the number of dermatologists will soon be insufficient to cope with the increased workload. New ideas are sorely needed.
Wanted: innovative solutions
I ask often myself why around 30% of melanoma are still detected at an advance stage? Why are there so many benign lesions excised – about 20 to 30 times more than the number of malignant lesions?
Why is it so difficult for general practitioners, who are very often acting as a first point of contact, to do an efficient skin exam?
I believe that with new technologies and e-health, which allow us to set up new processes and bring innovative healthcare services to clinics, solutions can be found to overcome those issues.
When dermatologists began using dermoscopy in the 1990s (or epiluminescence microscopy) it improved doctors’ capacity to see deeper under the skin. This enhanced specificity and sensitivity of diagnosis by up to 35% compared to the naked eye.
But, while the test itself takes just a few seconds, consultations with dermatologists take an average of 15 to 30 minutes. This limits the number of people who can access specialist skin check-ups. Others must rely on general practitioners who act as ‘gatekeepers’.
What if there is another way that would improve access while optimising the use of specialist time?
At Dermosafe we have built an IT platform that does exactly this. By arming healthcare assistants with a simple, user-friendly point-and-click device, we can take high-resolution images of patients without requiring a specialist consultation. These images are then viewed through a highly-secure web platform where they can be analysed by specialists trained for digital dermoscopy.
They system is supported by an artificial intelligence (AI) program that ensures the highest level of quality and performance. The result is a technology that increases the screening capacity of specialists by a factor of 10.
The tool has already been used to test around 15,000 moles on patients in France and Switzerland.
This blog is part of the MedTech Forum blog series. Philippe Held was a speaker at the MedTech Forum 2016. You can follow the conversation under #MTF2016 and find more details and at medtechforum.eu.