Better prediction of heart failure with data and computer models

Ventricular tachycardia (VT or V-tach) is a type of abnormal heart rhythm. A lot of data, among others, echo, MRI, ECG and lab values, of patients suffering from this condition is available in the hospital, which isn’t used at the moment. In a collaboration between de Eindhoven University of Technology (TU/e), Philips Research, and the Catharina Hospital, the COMBAT- VT project aims to develop a computational model that makes use of the patient’s data to predict the risk of cardio arrhythmias and supports the decision of the best treatment. This would significantly improve quality of life while reducing health care costs, says TU/e in a press release.

Patient data will be used for treatment, but there is also physiological and physical knowledge available in the form of computer models. The project will look into Physics Informed Neural Networks (PINNs), which tries to train on the data while also respecting the laws of physics. In this way training becomes more efficient.

AI system predicts risk of heart attack based on retinal scan

Is the eye a window into heart disease? An international team of researchers has developed an AI system that can identify patients who are in risk of a heart attack.

Electro-mechanical modeling

Electrophysiological VT simulations rely on geometrical information of the infarct area, where the remodeling of tissue properties is limited to this area in the current electrophysiological modeling pipeline. However, in mechanical simulations, functional abnormalities are detected outside this infarct area as well, where deformations of the nearby remote area are affected by the dysfunctional infarct.

In this project, such mechanical abnormalities are linked to electrophysiological remodeling. Here, abnormal strain distributions over the heart are assumed to affect the redistribution of remodeled electrophysiological properties over time. The effects of this hypothesis on long-term computational VT risk predictions are further explored to investigate if it might explain the progression of VT risk over time as observed in the clinic.

Reduce the computational effort

The new project also wants to make use of a new computational model for the analysis of VT data, named Isogeometric Analysis (IGA). IGA enables the construction of smooth discretized geometries. It is expected to reduce the computational effort when compared to the currently used methods while enabling an accurate representation of the development and behavior of VTs.