AI Can Now Diagnose Heart Disease and Lung Cancer

January 8, 2018 - 4 minutes read

Artificial intelligence (AI) has already made a multitude of contributions to MedTech in its short lifespan so far. Recently, Dogtown Media News explored how AI development is helping doctors with diagnoses as well as detecting chronic obstructive pulmonary disease (COPD).

Now, AI is becoming capable of diagnosing heart disease and lung cancer from scans. And it may start being used by the United Kingdom National Health Service later this year.

Improving Heart Disease Diagnosis

Conventional means to diagnose heart problems revolve around cardiologists recording the heartbeat rhythm through scans. 20 percent of heart scans are misdiagnosed every year, resulting in numerous patients receiving unnecessary surgery or falling victim to a heart attack. Annual expenses amount to £600 million (roughly $800 million).

Just an hour shy of London, developers at the John Radcliffe Hospital in Oxford have created an AI system that could dramatically reduce these figures by diagnosing heart disease more accurately than doctors. Called Ultromics, the system detects details in scans that human heart specialists can’t observe. Trained by reviewing the scans of 1,000 patients over seven years and utilized in the clinical trials of six cardiology units, the AI technology shows great promise according to Professor Paul Leeson.

Leeson, a cardiologist and the developer of Ultromics, discusses the need for AI in heart disease diagnosis: “As cardiologists, we accept that we don’t always get it right at the moment. But now there is a possibility that we may be able to do better.” The research results from the clinical trials will be published later this year and could be available in NHS hospitals as soon as this summer.

Other Innovation in Oxford

Heart disease diagnosis isn’t the only MedTech development in Oxford leveraging AI right now. Optellum is a local startup working on an AI system similar to Ultromics. Except instead of diagnosing heart disease, their AI focuses on detecting lung cancer by examining scans for suspicious cell clumps.

Also already being used in trials, Optellum appears to have plenty of potential. According to Timor Kadir, Optellum’s CSTO (chief science and technology officer), their AI system could possibly diagnose 4,000 lung cancer cases before doctors every year. This could result in cutting costs by as much as £10 billion ($13.5 billion) if used by both Europe and the United States.

Saving the NHS

Renowned geneticist Sir John Bell believes that AI may be the solution that the NHS needs to survive. “There is about £2.2 billion spent on pathology services in the NHS. You may be able to reduce that by 50 percent. AI may be the thing that saves the NHS.”

These savings could potentially be applied to other areas of need in the U.K. health system, like increasing the number of hospital staff and acquiring new equipment. Back in 2015, NHS England chairman Sir Malcolm Grant acknowledged AI’s ability to bring patients better care through improved diagnosis and customized treatment. While also noting the ethical challenges that AI could present, Grant urged the medical community to focus on the benefits, and that it would be “daft” to not utilize it.

AI’s role in MedTech will only grow with time. Although it will never quite take the role over from a human doctor (at least, not anytime soon), it has already proven to be an integral supporting tool in optimizing healthcare for patients. Expect to see its use expand into practically all areas of medicine in a relatively short time.

Tags: , , , , , , , , , , , , , , ,