Artificial Intelligence Topic of the Week

NumeriCor’s computer model can predict CRT success

Cardiac resynchronization therapy (CRT) is the most advanced piece of tech available for heart failure. The problem is that 30% of patients that undergo this invasive and expensive treatment don’t actually see any benefit. NumeriCor GmbH, a research spin-off based in Graz, Austria has developed a cardiac model that could predict who should get a CRT.

CRTs are good, but not for everyone

CRT is performed by implanting electrodes in the left and right ventricles of the heart, as well as on occasion the right atrium, to treat heart failure by coordinating the function of the left and right ventricles via a pacemaker, a small device inserted into the interior chest wall. CRT is indicated in patients suffering from a low ejection fraction (typically <35%) indicating heart failure with prolonged QRS duration.

Unfortunately, as many as 30% CRT recipients do not benefit from treatment (termed `non-responders’). The effort to improve patient selection in order to maximize human and financial resource utilization has fallen short so far. This is where advanced, ultra-personalized techniques may come in handy.

CARPentry-PRO modeling software

CARPentry-Pro is a cardiac simulation software developed by the Medical University of Graz and University of Bordeaux. It models cardiac electrophysiology, mechanics and hemodynamics at cellular, tissue and organ scales with high accuracy and performance. CARPentry-Pro and its academic predecessors have been used in dozens of peer-reviewed journal publications.

Super-accurate individual patient cardiac mapping is the key for CRT

The inter-university research network BioTechMed-Graz is partnering with NumeriCor to investigate whether cardiac modelling can predict CRT response. The researchers will use cardiac MRI imaging to map patients’ individual heart muscle fibers orientation, which is crucial in predicting the patterns of electrical signal propagation throughout the heart. With the help of AI algorithms, the researchers will be able to simulate the effect of bi-ventricular pacing, predicting how the heart will respond to CRT before even implanting the device! Better yet, this super-accurate cardiac model could guide implanting physicians to place the electrodes where they will optimally stimulate the heart. Gernot Plank, a Professor of Computational Cardiology in the Medical University of Graz believes that a clinical prototype will be operational sometime in 2022. The Fantastrial team will keep you posted!

These may also interest you

Apple Heart Study versus Huawei Heart Study

The Fantastrial Team

SentiAR: Electrophysiology meets AR


AIs that stare at ECGs: South Korean AI detects heart failure

The Fantastrial Team