Skin Analytics (SA) uses AI to give a diagnosis for cases of suspected skin cancer. It is positioned as a clinical grade tool for professional use. Performance is well evidenced to be similar to that of a dermatologist and far better than a GP.

Impact Objective:  Save lives – early detection of melanoma, as with all cancers, has a substantial impact on survival. Save money – Skin Analytics estimates that it can reduce the number of dermatology outpatient appointments by 50-60%, and costs by 50%.

Melanoma kills several hundred thousand people per year, but if caught early has a near 100% survival rate. There is a huge impact opportunity to create a better way to catch cancers earlier.

AI diagnostics provides such an opportunity, allowing ubiquitous, cheap and reliable analysis of images to give a strong indication of whether a skin problem is likely to be malignant.

What is required is to show that the technology works to the extremely high standards required for clinical application. Consequently Skin Analytics has undertaken one of the few real world prospective studies of the efficacy of AI diagnostics, and shown performance absolutely comparable to that of dermatologists. It is part way through FDA registration as a class III medical device, and has a number of further clinical studies in the works.

As well as saving lives, AI diagnostics should save huge amounts of clinical time. Of suspected Skin Cancers referred by GPs, more than 80% turn out not to be malignant. GPs are absolutely right to err on the side of caution here, but there is considerable room to provide tools that help them make accurate decisions.

We are big fans of Neil and his team at Skin Analytics, and believe they could have outstanding social impact in coming years.

© 2021 Nesta. Nesta is a registered charity in England and Wales 1144091 and Scotland SC042833. Our main address is 58 Victoria Embankment, London, EC4Y 0DS.
You can reach us by phone on 020 7438 2500 or drop us a line at information@nesta.org.uk.
All our work is licensed under a Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International License, unless it says otherwise. We hope you find it useful.