UCL AI Festival: the spinout ventures transforming patient diagnostics and healthcare with AI

UCL Innovation & Enterprise joined forces with NVIDIA and HP Enterprise for the four-day UCL AI Festival in central London, which kicked off 28 February. It featured panel discussions, hackathons, and presentations from thought leaders including Synthesia, and showcased some of the best advancements in AI currently.

As the UCL AI Festival comes to a close, we reflect on the wave of UCLB spinout ventures leveraging AI to drive a radical shift in healthcare, from robot-enhanced surgery to faster, better diagnostics.

UCL, which recently revealed that an AI‑powered life science district could add £3.5bn to the UK economy, was also home to a breakthrough by Queen Square Analytics (QSA). As reported in The Guardian, the development highlights how UCLB’s AI‑powered ventures are building tools and approaches with the potential to transform patient lives worldwide.

AI uncovers new disease profiles

QSA used AI to bring clarity to a notoriously complex condition: multiple sclerosis (MS), via sNfL (an early signal of nerve damage). Their models, for the first time, identified a more aggressive form of the disease, paving the way for truly personalised treatment options for patients. The work, carried out with UCL and drugmaker Merck, was peer reviewed in the journal Brain.

QSA was founded by Dr. Arman Eshaghi alongside the UCL Queen Square MS Clinical Trials Office and the UCL Centre for Medical Image Computing.

Precision for safer surgery

Another standout innovator stemming from UCL research is Panda Surgical. By rethinking what surgeons can achieve in the operating theatre, the company is potentially allowing for more complex operations on delicate areas, including the brain and the spine.

Panda’s handheld robotic platform, designed for minimally invasive neurosurgery, uses AI-supported exchangeable / detachable wristed surgical instruments that stabilise a surgeon’s movements. The tools could give a crucial advantage in avoiding scarring and making neurosurgery more accurate and safer.

Panda Surgical was founded by Dr. Manios Dimitrakakis (UCL Medical Physics & Biomedical Engineering), Professor Danail Stoyanov (UCL Computer Science) and Professor Hani Marcus (UCL Queen Square Institute of Neurology.

Enacuity is also advancing AI-assisted surgery. Co-founded by Dr. Maria Leiloglou, an honorary research fellow at UCL Medical Physics & Biomedical Engineering and Imperial, the company is improving surgeons’ ability to understand what is happened internally during keyhole surgery.

Its AI-powered hyperspectral imaging gives surgeons sharper, real-time visual guidance during complex laparoscopic procedures, helping clinicians ‘see’ the oxygenation level of tissues. This reduces the risk of complications and supports smoother recoveries. Bowel surgery patients, for example, have a better chance of smooth recovery if their tissue is well oxygenated.

Diagnostics for inherited retinal disease

For patients with inherited retinal disease (IRD), a leading cause of blindness, diagnosis can be a lengthy and stressful journey. Eye2Gene aims to change that by revolutionising scan analysis.

Using AI to analyse retinal scans, the system rapidly predicts which gene may be affected, helping clinicians move more quickly towards appropriate treatment planning.

Founded by Associate Professor Nikolas Pontikos at the UCL Institute of Ophthalmology, Eye2Gene is designed to mimic the approach of expert clinicians. The system learns from unlabelled eye scans, predicts the likely IRD, and is corrected by specialists, repeatedly refining its accuracy until it can reliably identify specific diseases.

AI for better diagnosis and recovery

Lastly, AI is helping healthcare teams improve diagnostics and speed of service by making sense of the huge amount of information stored in patient records. CogStack, a spinout venture created by UCL, King’s College London and several leading NHS Trusts, is building an AI platform trained on millions of real patient records and refined using hundreds of thousands of clinician-labelled examples, with strong medical accuracy.

The system has already identified eligible patients for major research projects such as the 100,000 Genomes Project, helping clinicians search unstructured notes to find patients with suspected rare diseases who might otherwise have been missed. It has also found thousands of missing fracture clinic records in minutes and reduced the time needed for medication reviews.

These examples represent just a handful of the ways UCLB spinout ventures are harnessing AI. The UCL ecosystem is working collaboratively to connect data better than before, helping the healthcare sector see things in new ways (sometimes, literally).

For us, the AI revolution is a chance to change things for the better.