Connect with us


Patient data reveals confidence in AI for detecting skin cancer



NHS patients with suspected skin cancer would rather be assessed by AI than wait weeks for in-person appointments, research from a UK med tech company has found.

Skin Analytics has published findings from a new research paper on its AI medical device DERM, which to date has assessed more than 53,000 patients for skin cancer in the NHS and correctly identified nearly 5,000 skin cancers across the UK.

The research examines the perspectives of 268 patients at Chelsea and Westminster Hospital which is currently using AI to identify skin cancer and support early diagnosis and treatment.

Its findings are overwhelmingly positive, with 62 per cent of patients saying they would rather have their skin assessed by a computer than wait weeks to see a dermatologist in-person.

More than two-thirds (69 per cent) also agree that having a computer assess photographs of their lesion saves time in comparison to a face to face consultation.

Meanwhile, 84 per cent felt comfortable having their lesions photographed with a mobile phone device, while 77 per cent have confidence that a computer can help them and their doctor by analysing photos of lesions.

Similarly, 84 per cent say having computers assess their photographs to help guide their GP is a good way of dealing with their problem.

Just 15 per cent say the prospect of having their lesions assessed by a computer made them feel uncomfortable.

The data reflects the increasing comfort levels among patients for virtual assessments, or teledermatology, conducted by both clinicians and AI, especially if it means expediting appointment times to be diagnosed, treated, or safely discharged.

Neil Daly, CEO of Skin Analytics, said: “Skin cancer accounts for the highest number of urgent referrals among all cancers in the UK yet, despite sustained and considerable effort from dermatology departments, waitlists are growing significantly.

“AI can both expedite and improve the accuracy of skin cancer diagnosis so that fewer unnecessary referrals are made, reducing waiting times and treatment delays for people with suspected cancer.”

The research comes as NHS England last month published its roadmap to accelerate the roll out of teledermatology, integrated with AI, to help manage demand and reduce unnecessary outpatient attendances.

There are around one million dermatology referrals in England every year, half of which are urgent suspected skin cancer referrals (previously known as the two-week wait skin cancer pathway).

Many of those referred on this pathway have benign skin lesions, but nationally 6.5-8.5 per cent will be diagnosed with melanoma and squamous cell carcinoma (SCC), the cancers these pathways are set up to catch.

Teledermatology can reduce the need to see people with benign skin lesions face-to-face and prioritise individuals with skin cancer.

In a separate analysis of DERM’s assessment of more than 53,000 NHS patients, 42 per cent of potential “two-week-wait” cases were deemed suitable for discharge – which could save as much as 5,700+ hours of dermatologist time.

The analytics shows AI’s enormous potential to triage patients into the right place, at the right time, and conserve NHS capacity for patients who need it.

Daly said: “Our latest evidence on DERM shows that patients’ principal concern is around the speed of assessment in diagnosing skin cancer, and they feel comfortable in choosing AI and teledermatology over traditional appointments should they be faster.

“In our latest quarterly performance review, we have seen our AI medical device find >95 per cent of skin cancers while discharging more than 7 in 10 benign lesions.

This is incredibly reassuring when we consider that survival rates for skin cancer soar above 95 per cent if detected early.

“Not only that, it reduces anxieties for those patients with benign cases who are waiting to be seen.” 

Continue Reading
Click to comment

Leave a Reply

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

Trending stories