Foot-recognition AI scanner can prevent heart failure hospitalisations

An AI device with ‘foot recognition’, similar to face-recognition technology, could potentially flag when heart failure is becoming severe and life-threatening, new research reveals.

© Aristal/Pixabay

© Aristal/Pixabay

The wall-mounted device, produced by a small Cambridge-based start-up company called Heartfelt Technologies, detects a build-up of fluid in the feet and ankles called oedema, which is one of the ‘big three' signs heart failure is becoming severe.

The device is typically installed beside someone's bed, where it scans their bare feet and lower leg. This is done automatically, and continuously, without requiring any involvement from the patient. The foot scanner simply needs to be plugged into the mains, and records real-time images automatically. It can work without Wi-Fi, for older people without internet access.

The scanner could deliver an alert 13 days before a person would end up in hospital, according to new research supported by the NIHR.

The device, which takes 1,800 pictures a minute, images the foot and lower leg at multiple angles using AI detection of their precise position, then calculates the volume of fluid they contain.

Dr Philip Keeling, senior author of the study and a consultant cardiologist at Torbay and South Devon NHS Foundation Trust, said: ‘This device detects one of the big three warning signs for people with heart failure before they end up in hospital.

‘Only about half of people admitted to hospital with heart failure currently get assigned an early review by a heart failure nurse who can check to see if they are suffering a harmful build-up of fluid because their heart is not working properly.

‘Amid a shortage of heart failure nurses, a device like this can be like a virtual nurse, tracking people's health.'

In The FOOT study the scanner accurately predicted five out of the six hospitalisations that occurred among participants. At the end of the six-month study, 82% of remaining study volunteers living at home – 18 out of 22 – chose to keep the device.

The researchers now plan to test the Heartfelt device, which can even scan feet through thin socks, on a larger cohort of patients.

So far it can accurately detect one person's feet and lower legs out of a group of 100 people's feet, but it may have identification issues if more people walk past, as would happen in larger residential homes.

North East and North Cumbria ICB launches medicines strategy

North East and North Cumbria ICB launches medicines strategy

By Liz Wells 05 June 2025

North East and North Cumbria ICB has unveiled a new medicines strategy, to cut down on unnecessary prescriptions and make the best use of NHS resources.

AI model generates personalised heart animations for better diagnosis

By Liz Wells 28 May 2025

Researchers at Imperial College London have developed a generative AI model that creates realistic and personalised animations of human hearts to help identi...

AI model to predict future health care needs

By Liz Wells 08 May 2025

Researchers at UCL and King’s College London are training an AI model that could identify opportunities where early interventions might significantly improve...


Popular articles by Liz Wells