Neil Daly, CEO of Skin Analytics: Helping detect skin cancer using a phone

AI skin cancer checks move from clinics to smartphones

Skin Analytics argues that AI can take on most of the work of finding skin cancer, because the great majority of people referred have nothing wrong. Years of NHS deployment and published results built the evidence, and a smartphone-only version now lets that capacity reach far more patients. The bet is that cheaper, faster triage catches cancers earlier and frees money for care.

Takeaways

• A smartphone-only tool is the key breakthrough. Launched in May with clearance across Europe to run autonomously without a clinician. It builds on the earlier dermoscopic version, which used a low-cost clip-on lens. Phones people already own become diagnostic tools, which changes access to care.
• Regulatory steps built trust from support to autonomy. Skin Analytics first won European clearance as a decision support tool for dermoscopic images, then reached an autonomous level where the AI decides without a clinician, then cleared the smartphone version. Six to seven years across the NHS built clinician trust.
• Results at scale back the technology. The company has seen more than 300,000 patients and found more than 25,000 cancers across the NHS. Daly cites a cancer pickup rate of around 97 to 98 percent and says performance data is published every quarter or six months.
• The economics favor avoiding dermatology appointments. About nine in ten patients on UK urgent pathways have nothing wrong, and only 6 to 8 percent of referrals are cancer. Appointments take fifteen to twenty minutes and are scarce. UK trusts spend around £80m a year on dermatology wait clearing.
• Early detection makes skin cancer ideal for AI. Skin cancer is visible on the skin surface before it spreads, and survival is about 98 to 99 percent when found at the earliest stage. Catching enough cases early requires seeing far more people than dermatologists can.
• Europe entry is complex and access must include care. Europe has one regulatory regime but many health systems with local requirements. Daly will not just publish an app, because patients need their next care step arranged, so the company works through providers across Europe.
• AI hype waves lifted belief in the product. Daly says each time AI regained public confidence, belief that people would accept AI diagnosis rose. Hospital adoption, previously slow, accelerated after ChatGPT. The company was early enough to make mistakes and learn the technology's limits.

What comes next

• Deploying with care providers across Europe is the immediate priority.
• Building the US business is also a priority, in a health care system that works very differently.