“AI in biotech” usually means drug discovery…especially recently.

The headlines go to the foundation models predicting protein structure, the generative chemistry pipelines spitting out novel molecules, the platform companies doing target ID.

The valuations are huge, the timelines are long, and the actual clinical impact remains TBD on most of it.

The quieter AI in biotech story, the one already saving patients, is surgical imaging.

Hypervision Surgical, a King’s College London spinout, closed a £17 million Series A led by Heal Capital, with Angelini Ventures, IP Group, and Daycrest joining. The cumulative raise is now around $21.5 million.

Proceeds are designated for commercial deployment, clinical adoption, and continued development of the company’s Hyperspectral Intelligence® platform.

I’m not SUPER familiar with the surgical tech industry but my guess is that $21.5 million will go much further here than in drug development.

So, what does this thing do?

How does it work?

What’s the deal?

Hyperspectral imaging captures light across many more spectral bands than a regular camera, which means it can pick up signal that’s invisible to the human eye, including subtle differences in oxygenation, perfusion, and tissue composition.

Hypervision’s system pairs that with AI-driven analytics to do this in real-time, wide-field tissue differentiation right in the OR.

It identifies tumor versus normal tissue, nerves, and blood vessels as the surgeon operates.

That’s obviously not a trivial problem.

Surgeons today rely on visible-spectrum imaging, palpation, and intraoperative pathology to decide what to cut and what to spare.

Margins on tumor resection are notoriously hard to call in real time.

Nerve preservation in cancer surgery is one of the highest-stakes judgment calls a surgeon makes.

If a hyperspectral camera plus a trained AI model can flag those tissues with quantitative confidence, the impact on outcomes is direct: fewer positive margins, fewer nerve injuries, less collateral damage.

Now, to be clear, Hypervision is not a Series A pre-revenue story.

They already have UKCA certification and FDA clearance for their first hyperspectral vision platform, with multi-center clinical evaluations underway in UK hospitals, primarily focused on gastrointestinal surgery.

GI is honestly a smart entry point.

Colorectal cancer surgery has well-documented margin issues, the procedural volume is enormous globally, and intraoperative tissue differentiation is something every GI surgeon would pay for if it works reliably in the OR.

The challenge ahead is clinical adoption.

UK hospitals are currently running evaluations. Real adoption requires NHS procurement decisions, US hospital purchasing, training programs for surgeons, integration with existing OR workflows, and reimbursement.

Each one is a slog.

The technology working is necessary but nowhere near sufficient.

If Hypervision pulls it off, this is one of those companies that quietly becomes part of standard surgical practice over the next decade, the way intraoperative ultrasound did, the way ICG fluorescence imaging is becoming.

Less spectacular than a lupus cure, but the kind of cumulative impact that matters.