Skin Analytics raises £4M Series A to use AI for skin cancer screening

Skin Analytics, a U.K.-based startup that has developed a skin cancer screening service that uses artificial intelligence, has raised £4 million in Series A funding. The round was led by Hoxton Ventures, with participation from Nesta and Mustard Seed Ventures.

Skin Analytics says it will use the injection of cash to expand its focus to the U.S. after it was awarded the “Breakthrough Device Designation” by the FDA as part of a programme designed to fast-track new technologies that can have significant impact on the nation’s health.

It also will continue forging partnerships within the U.K.’s national health service, following the launch of what it claims was the world’s first “AI-powered” clinical pathway in conjunction with University Hospital Birmingham.

Skin Analytics offers a CE marked medical device that studies suggests is able to identify skin cancers, pre-cancerous and benign lesions “to the same level as a dermatologist.” The idea is to enable health systems and insurers to increase dermatology capacity by reducing the burden of diagnosis for dermatologists.

“At its most simple, skin cancer is the world’s most common cancer and incidence is increasing around the world,” says Skin Analytics founder Neil Daly. “Overlay that with the fact there is a global shortage of dermatologists and we have a real challenge already with how we identify and deal with skin cancer.”

Daly says that Skin Analytics has developed a clinically validated AI system that can identify not only the important skin cancers, but the pre-cancerous lesions which can be treated by GPs and a range of benign lesions. “We can do that using a low-cost attachment and a smartphone, allowing us to put this into innovative patient pathways either at GP practices or in hospitals,” he says.

“By using our service, we can reduce the number of patients who end up in hospital by 40-60%, depending on where our technology is used. That [means]… we can reduce the demand on our scarce dermatology resources, freeing them up to focus on other patients such as the inflammatory skin disease patients who often wait months for appointments. We can also reduce the cost of skin cancer, freeing up that money to be reassigned to improving care elsewhere.”

Because skin cancer is such a large problem, coupled with advances in AI, Daly notes that there were initially many companies working in this space, seeing an explosion of competitors in 2014 with around 50 companies in the field. “All but three or four are gone now as the reality of how complex the technology is and how challenging operating as a clinical company hits home,” he says, before adding that there appears to be another wave of competitors surfacing.

“In reality, we’ve spent so much time learning from our mistakes in developing our AI, this is one of our main points of difference,” cautions the Skin Analytics founder. “It is too easy to get started and think you’ve made a great algorithm, but when you test it in the real world — and you can only do this with a prospective clinical study — the performance just isn’t there. Not only have we done that, but we use our research strategy to ask questions that give us the data to continue to improve our algorithms. There is no shortcut for this, you need to test, improve and repeat.”

Another key differentiator is that you can’t ‘fake it until you make it’ in a highly regulated industry and the processes that come with that. “You have to build them into the fabric of your company and it’s slow and painful,” adds Daly. “Medical device companies have to find a way to innovate quickly within a safety critical environment and I’m very proud of the way our team has built that ability, and continues to do so.”