The strongest positive AI story from the last 24 hours is a UK one, and it is potentially very significant for the NHS.
AI can spot hidden heart disease in seconds
Researchers at Imperial College London have developed AI that can extract far more information from a routine ECG than a human clinician can see unaided.
The system can flag signs of heart failure and valve disease from a standard ECG in under two seconds. In large-scale testing involving around 67,000 patients, reported accuracy reached roughly 81% for heart failure and 90% for valve disease.
An ECG is quick, cheap and already used everywhere. An echocardiogram is more expensive, takes longer and can involve significant waiting lists.
Rather than replacing cardiologists, the AI could act as an early-warning layer, helping doctors identify who needs further investigation most urgently.
The technology is now being trialled on 590 NHS patients in London and Bristol.
Source: Imperial College London — CardioVolt AI turns heart traces into powerful diagnostic tools ↗
Routine medical tests could become much more powerful
Researchers are increasingly treating the ECG as a digital biomarker, where AI can detect patterns linked not only to heart rhythm but potentially heart failure, valve disease, diabetes, kidney disease and other conditions.
That means a ten-second test the NHS already performs could eventually provide several layers of extra information without requiring entirely new infrastructure.
AI is also moving into blood-pressure treatment
Researchers have presented randomised trials examining AI-enhanced apps and wearables for managing hypertension.
These systems are being studied to see whether AI-assisted coaching, monitoring and personalised support can help people control blood pressure more effectively between GP appointments.
Why this matters
One of the most promising directions for healthcare AI is not machines independently diagnosing everything. It is turning ordinary tools such as ECGs, phones and wearables into much richer sources of medical information, then helping clinicians focus attention where it is needed most.
For the NHS, that could eventually mean earlier diagnosis, fewer unnecessary specialist tests, shorter waiting lists, better use of clinicians’ time and more monitoring at home.
Less hype. More real-world AI progress.