Connect with us


“AI in dermatology isn’t something that’s going to happen, it’s happening now”

Avatar photo



skin analytics

Skin Analytics is rolling out its AI-powered dermatology tool in NHS Trusts across the country. Health TechWorld talked to the company CEO and co-founder, Neil Daly, about what this means for AI in dermatology. 

Around the world, there is a shortage of dermatologists. In the UK the percentage of unfilled posts is approximately 30 per cent according to the latest figures. In the US, that number is as high as 50 per cent. With dermatology training taking upwards of a decade and training spaces limited, filling these posts will not be an easy task.

Meanwhile, skin cancer incidences are rising. According to Royal Marsden NHS Foundation Trust, the number of people diagnosed with melanoma has increased by almost half over the past decade. With a growing gap between supply and demand, artificial intelligence tools are being looked to as the solution.

Skin Analytics, a British med-tech company, aims to help more people identify skin cancer earlier so that it can be treated more effectively. Since 2012, Skin Analytics has worked with leading experts in dermatology, clinical research and general practice to develop DERM, which analyses dermoscopic images of skin lesions using AI algorithms.

The CE marked medical device is designed to relieve pressure on primary and secondary care services while helping more people survive skin cancer.

“The healthcare system has been run based on efficiency for a really long time now. The NHS is the world’s most efficient healthcare service and that means that we think about the world from a health economics perspective; how do we leverage technology to create value and save money for the healthcare system. And that’s a guiding principle of Skin Analytics.”

DERM is the only AI solution that can identify eleven different lesion types, including melanoma, the lesion that has the highest mortality rate, and non-melanoma skin cancers, the most common cancer in the world.

After capturing the potentially cancerous lesion with a smartphone or digital camera, the dermoscopic image is uploaded to the DERM cloud where it runs through a series of tests to ensure the quality of the image before being inputted into an algorithm that suggests what the lesion could be.

Operating alongside the medical device is a web-based teledermatology product that passes the information from DERM to the clinician and allows healthcare providers to build out different patient pathways.

“A lot of the information that we collect in the teledermatology platform is ultimately for the benefit of clinicians on the other end as well as the AI,” Daly explained.

“All of that information gets captured in this teledermatology platform, then DERM provides a classification and then this information comes back to the clinician with all of the medical history information. It helps them make a decision about what to do with that patient.

“They’re very much complementary products,” he added.

Skin Analytics was launched by Daly, a former strategy consultant, in 2012. According to Daly, the company was ahead of its time and its ambitions exceeded what was possible with the technology of the moment.

“I decided to start my own business and really try and bring some of the standardisation that computing power was starting to show was possible into healthcare and into diagnostics of skin cancer,” Daly said.

“It turned out we were quite far ahead of the technology, but then the wave of deep learning and convolutional neural networks exploded in 2013-2014 and that really transformed our ambitions and what was possible with the technology.”

Skin Analytics says it uses its AI to help increase dermatology capacity, speed up patient access to care, and decrease backlogs. The AI-powered tool allows patients with benign skin lesions to be identified earlier in their journey, directing dermatologists away from spending time diagnosing benign lesions so clinicians can spend more time with patients who really need their care.

“The idea is that we can take this technology, we can put it into hospitals, we can put it into primary care and eventually we can get closer to customers or patients [using it] in their own home. It means we can reduce the number of patients who end up in the hospital, because we can weed out the people who don’t need to from those that really do much faster and much more efficiently.”

Last month, Skin Analytics was granted a Phase 4 AI in Health and Care Award by NHSX, the highest phase in the competition.

Created to accelerate the evaluation of AI technologies within the NHS, the award is highly competitive with an extensive evaluation process. Receiving the award is a major validation of Skin Analytics’ DERM product and changes the way in which skin cancer is assessed and treated by the NHS.

DERM is already deployed in Mid and South Essex Health and Care Partnership and University Hospitals Birmingham (UHB), where, according to Daly, is the world’s first AI skincare pathway.

UHB was the first to adopt the platform, working with Skin Analytics to open an off-site facility for dermatological tests during the pandemic. The implementation of DERM has led to a reduction in unnecessary referrals while also helping to tackle the dermatology backlog created by COVID-19. At the beginning of the project, the trust had a backlog of around 750 patients. That number is now down to a few dozen.

In Mid and South Essex, Skin Analytics is now in 190 GP practices. When deployed in primary care, Skin Analytics is used as an adjunctive tool used to aid GPs with their decision making and help them place patients on the right pathway from the start.

Now, through the award, the platform is being made widely available across the NHS through a series of six deployments across primary and secondary care that will give thousands of clinicians and millions of patients access to Skin Analytics. The company also has four other trusts in the pipeline where it will be deploying DERM independently from the NHSX award.

“This is a hugely important step,” Daly said. “There’s a lot of noise; there’s a lot of people playing in this space, and for us as a company, one of the challenges is trying to really help our partners understand the difference between a research project and a fully commercial validated service like we are.

“I think this [award] is going to really help us do that. It really will help us get a weight of evidence as to what it looks like in the real world when you apply artificial intelligence and I think that’s missing across almost every sector within AI in healthcare at the moment.

“We think it’s going to be the real fuel to the fire of AI adoption.”

One NHS site that Skin Analytics is working with is Chelsea & Westminster Hospital. Dr Lucy Thomas, a consultant dermatologist at the hospital said the company’s award is “great news for the NHS and dermatology departments”.

“It will allow us to gather real-world data to demonstrate the benefits of AI on patient pathways and workforce challenges,” she said. “Like many services, dermatology has severe backlogs due to the COVID-19 pandemic. This award couldn’t have come at a better time to aid recovery and give us more time with the patients most in need of our help.”

Although it has been “slow at first” getting NHS trusts on board with the technology. Daly said the momentum is building as the company builds evidence of the platform working in the real world. Over the next three years, Daly believes that the NHS will have adopted a “huge amount” of AI into its workflows.

“As an innovator, I’ve been involved in talking about and talking with the NHS for a number of years,” he added. “Often you go to these sessions and people very quickly get frustrated about how hard it is to work within the NHS and how slow it is to adopt and innovate.

“I don’t think that’s true because the NHS is huge and it’s got so many priority areas that it is working on, but when it decides to focus on an area, it moves really really quickly.”

Daly stressed that AI would never become a replacement for doctors, however he said there is a “huge opportunity” to change the way healthcare is delivered to make it more patient-centric.

He added: “Artificial intelligence in dermatology is not something that is going to happen, it’s something that is happening.

“There’s a group of AI healthcare companies that are really investing heavily and making sure that we put patient safety at the heart of the way that we deliver these services, but this is out there happening now and we want to make sure that people know that and embrace it.”

Continue Reading
Click to comment

Leave a Reply

Your email address will not be published. Required fields are marked *

Trending stories