Abstract
2D echocardiography (2DE) is the most widely used non-invasive cardiac imaging modality due to its availability and low cost. Using 2DE, cardiac function is assessed by combining information from different acquisition planes. However, these plane positions are unknown and therefore measurements of left ventricular (LV) mass, volumes, and ejection fraction rely on geometric assumptions. While cardiac magnetic resonance imaging (MRI) is the gold standard in assessing cardiac function, it is expensive and limited to major centres. To enable accurate LV assessment from 2DE, we trained a deep learning (DL) model to predict the position of 2DE views in 3D space and present a full pipeline for automated 3D LV modelling and quantification using registered 2DE acquisitions. Good view positioning of apical four-chamber and apical two-chamber views was obtained for 34 out of 38 subjects (89% feasibility). A comparison between LV volumes and masses derived from our model-based analysis of 2DE against gold standard MRI showed no significant differences in end-diastolic (ED) volume, end-systolic (ES) volume and ED mass, whereas significant underestimates of volumes and overestimates of masses were found using the guideline standard Simpson’s Biplane Rule. Automated positioning of 2DE views in 3D space is feasible and important to overcome current geometrical assumptions in LV mass and volume estimation. This approach is readily available for clinical evaluation and applications involving cardiac digital twins.
Links to Paper and Supplementary Materials
Main Paper (Open Access Version): https://papers.miccai.org/miccai-2026-sat/paper/STACOM2026_025.pdf
SharedIt Link: Not yet available
SpringerLink (DOI): Not yet available
Supplementary Material: Not Submitted
Link to Open Review
Open Review Page: https://openreview.net/profile?id=~Vera_H._J._van_Hal1
BibTex
@InProceedings{vanVer_3D_MICCAISAT2026,
author = { van Hal, Vera H.J. AND Zhao, Debbie AND Dillon, Joshua R. AND Quill, Gina AND Edwards, Nicola C. AND Doughty, Robert N. AND Young, Alistair A. AND Nash, Martyn P.},
title = { { 3D Left Ventricular Modelling and Quantification from 2D Echocardiography Using Automated View Positioning } },
booktitle = {Medical Image Computing and Computer Assisted Intervention -- MICCAI 2026 Workshops and Challenges},
year = {2026},
publisher = {Springer Nature Switzerland},
volume = {LNCS 17267},
month = {pending},
page = {pending}
}