AI vs heart disease: NHS and Heartflow partner on cardiac diagnostics

Chloe Kent 16 February 2021 (Last Updated February 16th, 2021 12:41)

NHS England has mandated that hospitals adopt the artificial intelligence (AI) powered HeartFlow platform to diagnose coronary heart disease. HeartFlow’s analysis could help reduce reliance on invasive diagnostic procedures and reduce hospital waiting times. Medical Device Network takes a closer look.

AI vs heart disease: NHS and Heartflow partner on cardiac diagnostics
Like most medical AI, HeartFlow is designed to streamline the day-to-day work of trained clinicians, rather than replace them. Credit: HeartFlow

A number of different diagnostic tests are employed in hospitals to help confirm a diagnosis of coronary heart disease (CHD), and to monitor the progression of the condition. A lot of these diagnostics, like blood tests, are fairly benign. However, some patients will have to undergo procedures – including angiograms and MRI scans – that are invasive, time-consuming, expensive or all three.

England’s NHS and NHS Improvement have now mandated that English hospitals start to use the artificial intelligence (AI) powered HeartFlow FFRct Analysis platform to diagnose coronary heart disease as part of its new MedTech Funding Mandate.

The mandate, which will begin on 1 April this year, aims to provide innovative medical devices and digital procedures to patients faster, improving patient care and reducing costs for the service.

HeartFlow chief commercial officer Lance Scott says: “Our hope is that this will be the driving force allowing the HeartFlow Analysis to become the new standard when it comes to diagnosing and treating heart disease.”

HeartFlow has also received extended funding through the NHS Innovation and Technology Payment (ITP) programme for the third year running. The programme seeks to accelerate the uptake of innovations in the health service by removing financial barriers.

FFRct helps clinicians find coronary blockages

The AI platform takes data from a patient’s coronary computed tomography angiography (CTA) scan and uses deep learning technology and the input of trained analysts to create a personalised, digital 3D model of the patient’s heart.

Scott says: “The HeartFlow Analysis then uses powerful computer algorithms to solve millions of complex equations to simulate blood flow and provides FFRct [fractional flow reserve derived from CT] values along the coronary arteries.

“This information helps physicians evaluate the impact a blockage may be having on blood flow and determine the optimal course of treatment for each patient. A positive FFRct value (≤0.80) indicates that a coronary blockage is impeding blood flow to the heart muscle to a degree which may warrant invasive management.”

Using technology like this has several benefits when it comes to securing a diagnosis.

Newcastle Hospitals NHS Foundation Trust consultant cardiothoracic radiologist Dr Anna Beattie says: “The HeartFlow Analysis streamlines the diagnostic experience for patients. It significantly

reduces subsequent invasive diagnostic procedures for those who do not need revascularisation, achieving a management plan with a single non-invasive test.

“We have used the HeartFlow Analysis in Newcastle Hospitals since August 2018. Prior to adopting the technology, we used a coronary CT-first approach for 28% of all patients referred to our rapid access chest pain clinic. Now that has risen to more than 45% and almost 90% of patients going on to an invasive angiogram will get targeted revascularisation to improve anginal symptoms. Resources are better utilised by reducing diagnostic day case procedures and improving the efficiency of the catheterisation lab.”

HeartFlow’s technology is already being used in 60 NHS hospitals across England, but the extension of ITP and the introduction of the MedTech Funding Mandate will accelerate the adoption of the technology across the country.

Time well spent: AI to support clinician workflows

Like most medical AI, HeartFlow is designed to streamline the day-to-day work of trained clinicians, rather than replace them.

“The HeartFlow Analysis is a great tool to supplement a CT-first approach by providing functional information on top of the anatomical information,” says Beattie. “This helps reporting clinicians have greater confidence in their diagnosis and gives referring clinicians a more accurate indication of the significance of any atheroma detected and whether referral for revascularisation may be appropriate. It helps interventional cardiologists plan procedures to target functionally significant disease.”

For patients, the speedier pathways to diagnosis and reduced time spent in hospital have proven understandably popular. As the Covid-19 pandemic continues, technologies like HeartFlow’s are only likely to grow in popularity, as they limit the amount of face-to-face contact needed with doctors.

Beattie says: “For the patient, the technology limits the number of investigations required to make a diagnosis, reduces the time spent in hospital and face-to-face clinical contact. Our patient satisfaction survey has demonstrated the experience of undergoing cardiac CT to be very positive. Streamlining hospital visits is particularly crucial during the Covid-19 pandemic.”