EuroHPC-prosjekt MICROCARD: Numerical modeling of cardiac electrophysiology at the cellular scale

Cardiovascular diseases constitute the most frequent cause of death worldwide and half of these deaths are due to cardiac arrhythmia, which are disorders of the heart's electrical synchronization system. Mathematical models supported by computer simulations are essential to understand the behaviour of this complex system and its diseases. These models are already very sophisticated and widely used, but currently they are not powerful enough to study all the heart's (2 billion!) individual cells. It is therefore assumed that hundreds of cells are doing approximately the same thing. Due to this approximation, current models for example cannot represent the events in ageing and structurally diseased hearts, in which reduced electrical coupling leads to large differences in behaviour between neighbouring cells, with possibly fatal consequences. To model the heart cell by cell, we face a mathematical problem that is about 10,000 times larger, and also harder to solve. We will need larger supercomputers than those that exist today, and a lot of inventiveness to solve our problem efficiently on these future machines. The purpose of the MICROCARD project, which consists of 11 European partners, is to develop software that can solve this problem on future exascale supercomputers. We will also develop algorithms that are tailored to the specific mathematical problem, to the size of the computations, and to the particular design of these future computers. The software and simulators developed are intended to solve real-life problems in cardiology. Therefore the project includes computer experts, mathematicians, and biomedical engineers, and collaborates with cardiologists and physiologists. During the project period, we have studied how to best tackle different aspects of this problem, such as the numerical scheme and algorithms to be used. We have also developed software for simulation and for mesh handling, as well as performing tests to see how our methods and software perform. Most of this work has been presented at scientific meetings and published in peer-reviewed scientific journals. A follow-up centre of excellence, named MICROCARD-2, will continue with the research activities started in the MICROCARD project.

Several software packages have been enhanced and extended due to MICROCARD research activities. These include openCARP (https://opencarp.org), mmg3d & ParMmg (http://www.mmgtools.org), and Ginkgo (https://github.com/ginkgo-project/ginkgo). The software packages can considerably simplify the adoption of HPC in developing biomedical simulators that have application paths in medicine. Moreover, an open repository containing the usecases of MICROCARD has been set up at https://sites.google.com/simula.no/microcard-use-cases/home. The examples there demonstrate how HPC can be used to improve research in biomedical applications. The main simulation software developed in the MICROCARD project is called µCARP. It is implemented as a branch of the open-source openCARP software (https://opencarp.org). The research activities of MICROCARD have also contributed to a new Jupyter front-end of openCARP, which can greatly simplify code access for both academic and commercial users. In connection with the code development in MICROCARD, a Ginkgo back-end is added to openCARP and µCARP. This allows linear system solvers (one of the most important parts in any cardiac electrophysiology simulator) to run efficiently on GPU, which is a more energy-efficient architecture than standard CPU. A proof-of-concept study has been carried out to port and optimise a cardiac electrophysiology simulator for the low-power intelligence processing units (IPUs) newly produced by the European vendor Graphcore. This demonstrates the possibility of adopting European low-power processor technology for simulations of cardiac electrophysiology. The composition of the MICROCARD participants covers an unusually wide spectrum, from experts in cardiology, applied mathematicians to researchers in the wide domain of HPC. Moreover, the participants are from both academia and industry. For Simula Research Laboratory, the Norwegian partner in MICROCARD, participation in the project has helped strengthening its research activities in computational electrocardiology, while also improving the annual summer school on computational physiology co-organized by Simula, University of Oslo and University of California, San Diego.

Cardiovascular diseases are the most frequent cause of death worldwide and half of these deaths are due to cardiac arrhythmia, a disorder of the heart's electrical synchronization system. Numerical models of this complex system are highly sophisticated and widely used, but to match observations in aging and diseased hearts they need to move from a continuum approach to a representation of individual cells and their interconnections. This implies a different, harder numerical problem and a 10,000-fold increase in problem size. Exascale computers will be needed to run such models. We propose to develop an exascale application platform for cardiac electrophysiology simulations that is usable for cell-by- cell simulations. The platform will be co-designed by HPC experts, numerical scientists, biomedical engineers, and biomedical scientists, from academia and industry. We will develop, in concert, numerical schemes suitable for exascale parallelism, problem-tailored linear-system solvers and preconditioners, and a compiler to translate high-level model descriptions into optimized, energy-efficient system code for heterogeneous computing systems. The code will be parallelized with a recently developed runtime system that is resilient to hardware failures and will use an energy-aware task placement strategy. The platform will be applied in real-life use cases with high impact in the biomedical domain and will showcase HPC in this area where it is painfully underused. It will be made accessible for a wide range of users both as code and through a web interface. We will further employ our HPC and biomedical expertise to accelerate the development of parallel segmentation and (re)meshing software, necessary to create the extremely large and complex meshes needed from available large volumes of microscopy data. The platform will be adaptable to similar biological systems such as nerves, and components of the platform will be reusable in a wide range of applications.