Skip to main content

Artificial intelligence could improve heart treatment

The Heart Foundation has granted researcher Dr Jichao Zhao a three-year fellowship to investigate ways that the latest modelling and digital simulation could be used to improve treatment for a common heart problem in New Zealand.

Artificial intelligence (AI) and virtual 3D hearts are being used for world-leading research into atrial fibrillation (AF), New Zealand’s most common heart rhythm condition.

University of Auckland Senior Research Fellow Dr Jichao Zhao hopes a three-year Heart Foundation-funded fellowship will help identify why some people with AF don’t respond to current treatment strategies.

More than 60,000 people in New Zealand are diagnosed with AF, and it can lead to serious complications like stroke or heart failure. It is one of the country’s leading causes of hospitalisation for people with heart problems.

Initially AF may only occur occasionally and for short periods of time, but as the condition progresses episodes become longer and more frequent. Eventually, the irregular heart rhythm may become permanent.

Looking to improve treatment success rates

One current AF treatment is ablation, a procedure which treats areas of heart tissue that are causing the heart rhythm problem. However, for many people this treatment loses its efficacy over time.

“For patients with persistent or permanent AF, ablation is only successful in less than 30% of people at five years post procedure,” Jichao explains. “That’s a pretty low success rate, and that’s what our research is looking at. We’re trying to improve the ablation success rate for those people with a stubborn type of AF.”;

To do this, the team is focusing their efforts on the right chamber of the heart, the right atrium. Currently most research relating to AF, and the majority of ablations, target the left chamber.

The research will use computer modelling and artificial intelligence, to better define the muscular fibre networks that cause AF in the right chamber. It will look at how these fibres change in people with AF, and how those differences might contribute to treatment failure.

“Tissue analysis will be conducted using artificial intelligence. In current clinical practice, the medical images, including magnetic resonance imaging (MRI), are eyeball-checked to make decisions, which is subjective and prone to errors,” Jichao says.

“3D virtual hearts are not widely used in clinics to guide treatment. We will develop a robust, automatic clinical software program for creating 3D reconstructions of heart chambers to guide diagnosis, disease monitoring, treatment planning and prognosis of AF.”

Better outcomes without increased risk

The team will also try to determine the most effective ablation procedures for isolating right-sided AF, using digitally simulated ablation strategies in 3D virtual hearts.

“In this way, they can test the best ablation strategy for each patient without any additional risk to those patients,” Jichao adds.

Ultimately, he hopes this world-leading approach using computer simulation and artificial intelligence will lead to better clinical outcomes.

“The idea is that we improve our understanding of AF in the right atrium and come up with a better procedure to improve ablations for the patient with AF.”

Gaining better understanding of AF in Māori

The team also plans to focus some of the research specifically on the Māori population, who have a greater incidence of AF and with an earlier onset. To understand the reason for this disparity, his team will study the difference in the right atrial tissue between New Zealand European and Māori patients

Jichao has been involved in heart research for 14 years and heads up a team of 10 at Auckland Bioengineering Institute, at the University of Auckland, whose research focus is AF. As a mathematician by background, the use of mathematics, artificial intelligence and computer simulation has an exciting future in the clinical context.

“There will definitely be greater use of it in the future. In this context, we hope that we can use it to guide cardiologists to be able to perform a more targeted and accurate treatment.”

Read more research stories