Cardiovolt.ai raises £1.4M to turn ECGs into superhuman diagnostic tools
What's the deal? Cardiovolt.aiDealroom has a profile for this one. Try Dealroom →, a spinout from Imperial College London's National Heart and Lung Institute (NHLI), has raised £1.4 million to commercialise AI models that extract far more diagnostic information from standard electrocardiograms (ECGs) than any human cardiologist can.
The company's technology reads a routine ten-second ECG trace and can diagnose hidden heart conditions, flag non-cardiac diseases like diabetes and kidney disease, and predict a patient's risk of death. For heart disease, diagnostic accuracy reaches 83-93%; for non-cardiovascular conditions, 70-80%.
"We looked at the ECGs to see if we could do things that are superhuman," said Dr Arunashis Sau, Cardiovolt.ai's chief scientific officer. "Not things that clinicians can already do, but things that no cardiologist, no matter how expert, can do."
The founding team includes Professor Fu Siong Ng as chief medical officer, Dr Libor Pastika as chief technology officer, and Boroumand Zeidaabadi.
Why now? The models were trained on millions of ECGs sourced from research collections in Brazil and the US, each linked to patient medical histories. That scale of data — needed for AI to learn effectively — has only recently become accessible.
The team also used the UK Biobank, where ECGs are paired with imaging data, genetic profiles, and protein markers, to understand the biological mechanisms behind the AI's predictions. What it detects is essentially a digital biomarker: a signal that disease is underway, embedded in the heart's electrical trace.
What could go wrong? Medical AI faces steep regulatory hurdles before it can be deployed in hospitals. Convincing clinicians to trust a model that finds signals invisible to the human eye is another challenge — particularly when the AI flags conditions like diabetes from a heart trace, a connection that may feel counterintuitive.
Accuracy rates of 70-80% for non-cardiac conditions, while impressive for a ten-second test, also mean significant false-positive and false-negative rates that would need careful clinical management.
The signal: Cardiovolt.ai sits at the intersection of two accelerating trends: AI-driven diagnostics and the commercialisation of UK university research. Classified as early growth stage on Dealroom, the company is betting that reinterpreting existing, low-cost medical infrastructure — rather than building new hardware — offers a faster route to clinical adoption. With billions of ECGs already recorded worldwide each year, even modest diagnostic gains at scale could reshape screening economics for cardiovascular and non-cardiac diseases alike.
Read more: imperial.ac.uk
