Finite Element Modeling of the Infant Heart to Determine the Relationship between Single Ventricle Disease and the Seismocardiogram

Treatments for Single Ventricle Diseases (SVD) involve risky surgeries and can still result in hospital readmission, making continuous wearable monitoring systems to detect clinical deterioration and improve patient outcomes extremely valuable. The seismocardiogram (SCG), a non-invasively acquired cardiomechanical signal, contains features that provide critical heart mechanic insights and can be used in conjunction with existing wearable signal modalities to improve SVD patient outcomes. However, the exact relationship between SCG features and cardiac behavior has not been thoroughly studied. While modeling heart movements has increased SCG understanding, much of the previous work has only used canine and adult human hearts. The hearts of patients with SVD are not well understood. In this work, we created 3D models of the infant's heart and developed a modeling pipeline to compare the acquired SCG signals to those created by simulation models. The result indicates accurate segmentations of the aortic root, ventricles, and atria, and a Finite Element Modeling (FEM) analysis on a ventricle heart model is established. Modeling heart movement for SVD patients can provide a deeper understanding of SCG signals in identifying specific SCG characteristics in SVD patients for monitoring applications.