Page 27 - Artificial Intelligence for Computational Modeling of the Heart

xxviii Preface

iology and pathology. The reader will discover methods and ap-
proaches that combine the best of both sciences, for fast, accurate
and personalized digital twins of the heart.

Book organization

The book is organized into two parts. First, general cardiac
physiology, modeling, and implementation concepts are intro-
duced. The objective is to provide the reader with the necessary
background that will then be useful for the remainder of the book.
The second part describes how machine learning and artificial in-
telligence can facilitate patient-specific modeling, ranging from
automatic anatomical modeling to AI accelerated models to ro-
bust parameter estimation. Finally, the book concludes with clini-
cally relevant examples to illustrates how such technologies could
be applied to real-world applications.

Part 1: Modeling the beating heart: approaches and implementation
Chapter 1: Multi-scale models of the heart for patient-specific
simulations. The first chapter of the book introduces the main
concepts and models of heart function. Physiological mecha-
nisms and related models are introduced and discussed. The ma-
jor physiological systems are described: anatomy, electrophysiol-
ogy, hemodynamics and biomechanics. Methods used for param-
eter estimation from clinical data are also introduced.
Chapter 2: Implementation of a patient-specific cardiac model.
The second chapter deep dives in one exemplary model and
describes implementation strategies for each physiological sys-
tems. More precisely, lattice-Boltzmann methods are presented
for computational fluid dynamics and electrophysiology. Graph-
based approaches for real-time cardiac electrophysiology simula-
tions are also detailed. A computationally efficient finite element
method for modeling soft tissue deformations, based on the To-
tal Lagrangian Explicit Dynamics framework, is introduced for
the computation of bi-ventricular electromechanics. Finally, the
chapter closes with an integration strategy to efficiently perform
fluid-structure interaction simulations.

Part 2: Artificial intelligence methods for cardiac modeling
Chapter 3: Learning cardiac anatomy. The first step in patient-
specific modeling is the fast and robust parsing of the cardio-
vascular anatomy from medical images. This task usually in-
volves detecting, segmenting and tracking anatomical structures
or pathologies in the human heart. Current approaches are based