Oxford University spin-out uses AI to spot heart disease
May 19, 2020
Ultromics, a spin-out company from the University of Oxford, has developed a medical device that uses artificial intelligence (AI) to predict coronary artery disease.
Heart disease is the number one cause of death in the world and affects half of men and a third of women. It kills one in seven men and one in twelve women in the UK, striking every eight minutes.
The technology, called EchoGo Pro, uses AI to make predictions about coronary heart disease risk through ultrasound analysis, the primary imaging test to diagnose heart disease, with 60,000 stress echo scans performed in the UK each year.
EchoGo Pro has been CE marked allowing it to be used across the UK and the EU, with plans to submit for FDA clearance, so it can be used in the USA by 2021.
It can deliver heart disease prognosis to hospital sites through its secure cloud system, built in partnership with the UK NHS, based on thousands of past clinical exams. The system automatically analyses ultrasound images and applies AI to identify disease, analysing thousands of data points, compared with five or six indicators in traditional visual inspection. It generates a report in minutes and sends a prognosis to the doctor treating the patient.
"EchoGo Pro identifies the risk of heart disease, helping to detect early signs and enable preventative steps before a heart attack strikes, ultimately saving lives,” said Ultromics co-founder and CEO Ross Upton. "The high level of precision and its speed of use means doctors, regardless of their level of experience or training, can make accurate recommendations to patients to help reduce misdiagnosis."
Unfortunately, misdiagnosis of cardiovascular disease occurs in one in five patients due to a complex web of symptoms, circumstances and comorbidities, which doesn't make it easy for doctors to identify conditions correctly in traditional visual inspection. EchoGo Pro is said to be able to identify these hard to spot features.
"Misdiagnosis rates will be much lower and at-risk people can be treated quickly while those not at risk will be spared unnecessary tests, surgery and other treatment," said Upton.
The reduction in misdiagnoses can make a big difference to hospital and healthcare providers, saving them money and medical resources. It can save valuable time for doctors and the healthcare system. The AI system streamlines hospital's workflow to help reduce the burden on doctors, freeing up their time so they can spend time with patients and potentially treat more people.
The technology was trialled in the UK and USA and achieved a diagnostic performance of over 90% (AUROC), halving the number of misdiagnoses compared with reports of routine clinical practice. The technology has been developed through its partnership in the Evarest trial with the NHS, one of the largest ultrasound programmes in the world, trained on tens of thousands of heart scans. The Evarest validation trial continues to run in over 30 NHS hospitals and has recruited over 6000 patients.
"The CE mark for EchoGo Pro is an important time in the battle against heart disease where real progress is being made to help improve patient outcomes and prevent heart disease,” said Paul Leeson, head of the Oxford Cardiovascular Clinical Research Facility, who co-founded Ultromics with Upton. “Leveraging AI technology means we can more accurately predict heart disease and optimise care pathways to help make valuable cost and time savings for healthcare systems at this time when they are already stretched and in much need of support."
In time, Ultromics hopes to develop versions of EchoGo Pro that can be deployed in handheld devices, and which can be used to improve heart function testing and analysis in remote areas of the world, or places with limited health infrastructure.
EchoGo Pro is not for primary diagnosis of coronary artery disease, diagnosis of mild or moderate myocardial ischemia, or localisation of coronary artery disease or myocardial ischemia. It is not for the assessment of myocardial perfusion, myocardial viability or valve disease.
Founded in 2017, Ultromics applies AI to develop echo-based tools in cardiovascular imaging, by combining deep clinical insight with machine learning and some of the largest echo datasets in the world. The platform assists clinicians in their decision-making, without disrupting clinical workﬂow, to drive diagnostic quality and patient care.
Each year cardiovascular disease causes an estimated 17 million deaths and is the largest cause of death globally.