AI Colour-Codes the Body in Real Time During Surgery — and the NHS Just Ran Its First UK Trial

Surgeons at St Mark's Hospital in London have used an AI tool that maps and colour-codes human anatomy live during an operation — a UK first that puts machine vision inside the operating theatre. For NHS Scotland, which has been quietly building its digital health infrastructure, this is the kind of clinical AI that changes outcomes, not just workflows. Here's what happened, and why it matters north of the border.

The procedure took place at St Mark's Hospital in London, a specialist centre for bowel disease, where surgeons used an AI system capable of identifying and colour-coding anatomical structures in real time as the operation unfolded. The tool overlays the live surgical feed with a visual guide — think thermal imaging, but for tissue type — helping surgeons distinguish between structures that, to the naked eye under pressure, can look dangerously similar. It is the first time this technology has been used in a live operation in the UK.

The clinical stakes here are significant. Accidental damage to bile ducts, ureters, or major blood vessels accounts for some of the most serious surgical complications — injuries that are difficult to repair and costly in every sense. According to NHS England's Patient Safety data, bile duct injury alone affects roughly 1 in 300 to 1 in 500 laparoscopic cholecystectomies, with consequences that can be life-altering. AI that reduces that margin of error even fractionally translates directly into patient outcomes and reduced litigation costs for health boards.

The technology builds on a growing body of work in computer vision applied to surgery. Research published in journals including Nature Medicine has shown that AI models trained on thousands of surgical video frames can identify anatomical landmarks with a consistency that tired or pressured humans sometimes cannot. The AI does not replace the surgeon's judgement — it adds a layer of persistent, unemotional attention that never has a bad day, never gets fatigued at hour seven of a complex procedure.

For NHS Scotland, the question is now trajectory and timing. Scotland's Digital Health and Care Strategy, backed by the Scottish Government, explicitly commits to adopting clinical technologies that improve safety and efficiency. Healthcare Improvement Scotland (HIS) has a clear mandate to evaluate and roll out innovations of this type. Scotland's smaller, more integrated health system — compared to the fragmented NHS England structure — arguably makes it faster to adopt and scale surgical AI once the evidence base is established. A UK-first trial in London is, in practical terms, a proof of concept for Edinburgh, Glasgow, and Aberdeen.

What this also signals is the broader direction of travel for AI in medicine: not replacing clinicians, but dramatically amplifying what they can see, decide, and do. The surgeon at St Mark's still held the instruments. The AI held the map. That division of labour — human dexterity and judgement, machine-level pattern recognition and consistency — is the model that is going to define operating theatres within a decade. Scotland's health boards, procurement teams, and medical training institutions would do well to get familiar with what that model demands of them now, not after it becomes standard everywhere else first.

What this means for you: If you work in medical device supply, health tech, or professional training for clinical staff in Scotland, this is a procurement and partnership signal worth acting on now. NHS Scotland's Digital Health and Care Strategy creates a clear pathway for exactly this kind of surgical AI — meaning there will be evaluation rounds, supplier calls, and training contracts coming. Check what Healthcare Improvement Scotland (his.scot) is currently inviting tenders for, and if you're building anything adjacent to clinical AI, get your case study in front of the right people before this becomes a crowded space.