Fascination With Technology Preventing Heart Diseases with AI

Related Vendors

Fachhochschule München

FAULHABER Drive Systems

dataTec AG

Wuxi InfiMotion Technology Co., Ltd.

In our "Fascination with Technology" section, we present impressive research and development projects to engineers every week. Today: how artificial intelligence (AI) can help prevent heart diseases.

Cardiovascular diseases are the leading cause of death in Germany. More than 300,000 people die each year from cardiovascular conditions such as heart attacks, strokes, heart rhythm disorders, or high blood pressure. "The causes of these diseases are likely often multifactorial: there are numerous interactions, for instance, between blood pressure, the shape and function of the heart muscle, and the heart valves. These complex relationships are difficult to study in living patients," explains Ludwig Wagmüller.

The mechanical engineer developed a personalized computer model of a pulsating cardiovascular system during his doctoral work at the Munich University of Applied Sciences (HM), Germany. This is intended to make it possible in the future to analyze the heart's behavior without invasive diagnostic procedures.

Supercomputers required several hours to calculate and visualize a single heartbeat.

Ludwig Wagmüller, PhD student at Munich University of Applied Sciences

Previous Simulations Often Slow And Inefficient

Previous simulations were on the one hand too slow for this, and on the other hand, only adaptable to patients in a complex manner. "Supercomputers required several hours to calculate and visualize a single heartbeat," explains Wagmüller.

Together with simulation experts from the Faculty of Mechanical, Automotive, and Aeronautical Engineering at HM and the Technical University of Munich (TUM), he designed a novel heart model using AI methods: It can accurately replicate patient-specific geometry while requiring less computing power than traditional simulations.

We use a combination of statistical methods and AI. This approach ensures that the simulation requires less computing time.

Ludwig Wagmüller, PhD student at Munich University of Applied Sciences

Faster Computing With Artificial Intelligence

The trick: "We use a combination of statistical methods and AI. This approach ensures that the simulation requires less computing time," says the doctoral student. A key role is played by the "Reduced Order Model." Such reduced models are less complex than classical simulations but achieve high accuracy by considering essential characteristics and are also much more energy-efficient. For the first time, the researchers were able to identify and mathematically describe typical motion patterns in heart movement across different patient geometries.

The Digital Twin of the Average Heart

The new heart model is based on real data from living patients. Using seventy anonymized MRI datasets, Wagmüller succeeded in simulating the digital twin of an average heart, including its variations. This was then trained with additional anonymized MRI data.

The result is a pulsating digital cardiovascular system that can model and predict essential physical processes. This digital twin can be personalized using specific data.

By combining reduced models, which accelerate simulation, with variable geometries, which allow for customization, this opens up entirely new applications for simulation technology.

Markus Gitterle, Professor at Munich University of Applied Sciences

A Model on Its Way to Clinical Practice

"By combining reduced models, which accelerate simulation, with variable geometries that allow for customization, this opens up entirely new applications for simulation technology," summarizes HM Professor Markus Gitterle from the Faculty of Mechanical, Automotive, and Aeronautical Engineering, who led the project together with HM Professor Michael Wibmer.

In the long term, the digital twin enables insights into pulsating cardiovascular systems. A future vision of the researchers is the visualization and testing of surgical interventions: "The digital twin is continuously being developed. In this way, it may one day be possible to examine before open-heart surgery whether the planned operation will achieve the desired success," adds Wibmer.

The SmartHeart project is funded by the Bavarian State Ministry of Science and the Arts and implemented in collaboration with project partners AdjuCor GmbH and Prof. Dr.-Ing. Michael W. Gee of the Technical University of Munich.

Future-Proof Authentication with Universal RFID Readers

Universal RFID Readers: One Solution for All Credentials