VVZ API is not affiliated with ETH Zurich. Data might be outdated or incorrect. Please view the official ETHZ Vorlesungsverzeichnis for binding information.
Abstract
Der Bereich Praktika, Projekte, Seminare umfasst Lehrveranstaltungen in unterschiedlichen Formaten zum Erwerb von praktischen Kenntnissen und Fertigkeiten. Ausserdem soll selbstständiges Experimentieren und Gestalten gefördert, exploratives Lernen ermöglicht und die Methodik von Projektarbeiten vermittelt werden.
Objective
Digital twins are in-silico replica of physiological systems modelling anatomy and/or function. They are increasingly being used for data augmentation of partial data, biomarker calculation and evaluation of multiple functional conditions [1,2]. This course will provide hands-on experience on the generation of digital twins for cardiac applications. It will cover an introduction of basics concepts of cardiac biomechanics, numerical modelling and physics-informed neural networks [2,3]. In the practical part, the students will apply these concepts for the generation of a cardiac digital twin from Magnetic Resonance Imaging data. These models will be used to extract clinical indices (ejection fraction, strains, myocardial mass) and simulate physiological and pathological function. Learning objectives: Upon completion of the course students are able to: • understand the physiological basis of cardiac mechanics and function and the clinical indices associated with their assessment • understand the basics of neural networks and physics-informed neural networks • build anatomical digital twins of cardiac anatomy from patient-specific magnetic resonance images • personalize cardiac functional twins using physics-informed neural networks Prerequisites: • Python • English (the course will be thaught in English) References [1] Joyce, T et al. Rapid inference of personalised left-ventricular meshes by deformation-based differentiable mesh voxelization, Medical Image Analysis, 79, 102445 (2022) [2] Buoso, S., Joyce, T., and Kozerke, S. Personalising left-ventricular biophysical models of the heart using parametric physics-informed neural networks. Medical Image Analysis 71,102066 (2021) [3] Buoso, S., et al. MRXCAT2.0: Synthesis of realistic numerical phantoms by combining left-ventricular shape learning, biophysical simulations and tissue texture generation. Journal of Cardiovascular Magnetic Resonance 25, 25 (2023). Dates • Week 1 o 19th February: 14.00-16.00 o 20th February: 14.00-16.00 • Week 2 o 26th February: 14.00-16.00 o 27th February: 14.00-16.00 • Week 3 o 5th March: 14.00-16.00 o 6th March: 14.00-16.00 • Week 4 o 12th March: 14.00-16.00 o 13th March: 14.00-16.00 • Week 5 o 19th March: 14.00-16.00 o 20th March: 14.00-16.00 • Week 6 o 26th March: 14.00-16.00 o 27th March: 14.00-16.00 • Week 7 o 2nd April: 14.00-16.00 o 3rd April: 14.00-16.00