Session: Research Posters

Paper Number: 173937

Multiphysics Simulation of Scg Signals Using a Digital Twin of the Human Heart: Toward Virtual Diagnostics of Cardiovascular Diseases 

Mechanical dysfunctions of the heart, including congenital and acquired abnormalities, pose significant clinical challenges due to their impact on cardiac workload, hemodynamics, and the outcomes like heart induced chest vibrations known as SCG signals. Conditions such as valvular stenosis, regurgitation, and cardiomyopathies (whether dilated or hypertrophic), alter the mechanical behavior of the myocardium, often leading to chamber remodeling and reduced cardiac efficiency. In particular, septal defects like ventricular (VSD) and atrial (ASD) septal defects disrupt the normal pressure gradients between cardiac chambers and force the heart to work harder and potentially lead to pulmonary hypertension or heart failure. Similarly, coarctation of the aorta (CoA) and aortic stenosis (AS) impose pressure overload on the left ventricle. These mechanical anomalies, including congenital heart defects that affect nearly 1% of live births, carry significant risks. Mortality rates can exceed 60% in certain pediatric cardiomyopathies, and for untreated severe aortic stenosis, survival drops to below 10% within three years. As such, there is a critical need for early, non-invasive, and physiologically informed diagnostic tools that can improve the screening and understanding of these life-threatening conditions.

Digital twin models of the heart offer a transformative approach for simulating and understanding the mechanical effects of cardiovascular diseases by creating a patient-specific virtual replica of cardiac anatomy and function. These models integrate medical imaging, physiological measurements, and computational techniques to reproduce the structural and dynamic behavior of the heart in health and disease. For conditions such as valve stenosis, septal defects, and cardiomyopathies, digital twins can simulate how altered geometry, pressure, or compliance impacts blood flow and wall motion over the cardiac cycle. This capability allows researchers and clinicians to examine disease progression, predict mechanical consequences, and evaluate potential interventions without relying solely on invasive procedures or large clinical datasets. By simulating how specific abnormalities influence heart-induced chest vibrations, digital twins provide a valuable tool for correlating mechanical dysfunctions with external signals such as SCG. This contributes to a deeper understanding of how various pathologies affect SCG signals and supports the development of a non-invasive diagnostic method that is tailored to individual patients.

The methodology involved constructing a computational domain from 4D CT images of a healthy subject, encompassing the lungs, ribcage, chest muscles. The objective is to develop a complete digital twin model of the four-chamber heart and integrate it with the existing anatomical geometry. This model simulates the actual blood flow through the heart during a full cardiac cycle. The simulation includes the ejection of blood from the left and right ventricles into the aorta and pulmonary arteries during systole, as well as the filling of the ventricles from the atria during diastole. Since the primary source of SCG signals on the chest surface is the motion of the heart walls, accurately modeling this blood-induced motion provides deeper insight into SCG signal generation. By creating a patient-specific digital twin of the four-chamber heart and applying boundary conditions representative of a healthy individual, we aim to replicate physiological flow dynamics and the resulting mechanical interactions with surrounding tissues. The validation consists of two steps. First comparing the heart wall motions with the wall motions that are captured using Optical flow methods (3D Lucas Kanade) and also deep learning. Then the Multiphysics generated SCG signal compared with the FEM generated SCG signals (from our previous papers) and human measured SCG signals.  Once validated, this model can be modified by adjusting both boundary conditions and anatomical structures to simulate various pathological scenarios, offering a powerful framework for studying the mechanical signatures of cardiac diseases.

Preliminary results demonstrated the feasibility of the proposed methodology, including the successful simulation of blood flow dynamics within the four-chamber heart and the realistic opening and closing behavior of the cardiac valves using a coupled multiphysics framework. These simulations should finally produce coherent mechanical interactions between the heart and surrounding tissues that lead to SCG waveforms that closely resemble those observed in experimental and clinical measurements. This modeling approach offers a promising pathway for uncovering the biomechanical origins of SCG signals and may contribute to the development of reliable, non-invasive tools for screening and monitoring heart diseases. As a future direction, this digital twin framework can be extended to model specific pathological conditions, such as valve stenosis or septal defects, by altering the anatomical geometry or flow boundary conditions. This would enable personalized simulations of disease progression and intervention outcomes. It also highlights the growing role of computational modeling in clinical cardiology.

Presenting Author: Amirtahà Taebi Mississippi State University

Presenting Author Biography: Dr. Amirtahà Taebi is an Assistant Professor of Biomedical Engineering at Mississippi State University. He is a recipient of the National Science Foundation (NSF) CAREER award and was named an ASME Rising Star of Mechanical Engineering. He was also recognized with an honorable mention for the Skalak Award by the ASME Bioengineering Division. His work focuses on developing AI-powered sensing technologies, wearable and contactless monitoring systems, and digital twin modeling to advance noninvasive cardiovascular diagnostics. His research has been supported by local, state, and federal agencies, including the NSF, NIH, and Mississippi Institutions of Higher Learning.

Authors:

Mohammadali Monfared Mississippi State University
Peshala Thibbotuwawa Gamage Florida Institute of Technology
Bahram Kakavand Nemours Children's Hospital
Amirtahà Taebi Mississippi State University