Computational Modelling of the Heart: from Fundamentals to Clinical Applications
Recent progress in the fields of computational modeling, data science, and artificial intelligence have shaken the landscape of cardiovascular research. In parallel, state-of-the-art medical imaging technologies now provide streams of patient data of unprecedented volume and detail. The convergence of these advances with the growth in processing power and communication infrastructure has made it possible to run patient-specific models of high fidelity or in large numbers, and empowered inter-disciplinary collaborations that are accelerating further developments of multi-scale, multi-physics models.
These synergies are driving a revolution in cardiovascular science by offering increasingly realistic digital twins to investigate disease mechanisms, facilitate medical device design and implantation planning, search for therapeutic targets, and improve diagnosis. However, the translation of computational advances to improve clinical decision support systems also faces significant challenges. It is necessary to be able to handle and manage vast amounts of data, coming from various sources and with different formats. Running patient-specific models with high fidelity requires significant computational resources. Multi-scale and multi-fidelity models can alleviate this requirement but involve careful integration. Building end-to-end simulation pipelines demands thorough software engineering and multi-disciplinary collaboration. Acquiring the imaging data that feeds these pipelines is often challenging when it is not clinically indicated. Multi-physics models reduce the amount of necessary image data and allow for studying mechanistic questions. However, they rely on model parameters that are difficult to measure experimentally. Moreover, before clinical support systems can be implemented, rigorous validation against real-world clinical data is required. Other challenges are related to regulation, ethics or gaining the trust in the models of healthcare professionals, such as doctors and nurses, among others. Overcoming these challenges is essential to realize the full potential of current cardiovascular research and requires a new generation of scientists trained in all these aspects.
This course will review core competencies and novel developments in cardiac computational modeling. Particular attention will be given to multi-physics, multi-scale model development, digital twins, and physics-informed inference from medical images. We will consider patient-specific modeling beyond cardiovascular geometry, and review simulation pipeline automation and integration into clinical decision support systems. The course curriculum will cover fundamental topics in cardiac biomechanics and state-of-the-art modeling techniques. It will also discuss machine learning and AI tools for image analysis and provide examples of computational models adopted into clinical decision support systems.
The course is aimed at graduate students of applied sciences and engineering interested in all aspects of cardiac bio-electro-mechanics, including imaging and computing. Researchers specializing in one field (e.g., electrophysiology) interested in branching out to other areas (e.g., fluid-structure interaction) are also welcome to attend this course. The course will feature poster sessions to spark cross-disciplinary communication among participants.