Abstract
Cardiac statistical shape and motion atlases enable analysis of disease- related remodelling but are confounded by covariates such as body habitus. Here, we use shape-derived endophenotypes (SDEs) to quantify size, shape and motion variation associated with demographic factors, risk factors and disease status. Biventricular meshes were fitted to 76,885 UK Biobank CMR exams. Principal component multivariate regression yielded SDE scores in the direction of maxi- mum covariance with height, heart rate (HR), age and mean arterial pressure (MAP), using propensity-based balancing weights to isolate individual covariate effects. Removing height-, HR- and MAP-predicted components produced a con- founder-free shape space while retaining age-related variation. Within this space, an age SDE was associated with reduced diastolic relaxation strain rate, with preserved peak systolic strain but slower early diastolic recoil. As validation, a sex- stratified age SDE residual (shape-age gap) was constructed in both uncorrected and confounder-free shape spaces. The uncorrected-space gap showed spuriously stronger univariate associations with several cardiometabolic conditions, reflect- ing greater covariate-driven variance (R² = 0.212 vs. 0.016). After covariate correction, this advantage was largely removed: the confounder-free gap outper- formed the uncorrected-space gap across nearly all conditions in both sexes (ex- cept diabetes in women), demonstrating that covariate correction reduces confounding while preserving disease-relevant shape and motion information.
Links to Paper and Supplementary Materials
Main Paper (Open Access Version): https://papers.miccai.org/miccai-2026-sat/paper/STACOM2026_034.pdf
SharedIt Link: Not yet available
SpringerLink (DOI): Not yet available
Supplementary Material: https://papers.miccai.org/miccai-2026-sat/supp/STACOM_034_supp.zip
Link to Open Review
Open Review Page: https://openreview.net/profile?id=~Charlene_Mauger1
BibTex
@InProceedings{MauCha_Cardiac_MICCAISAT2026,
author = { Mauger, Charlène A. AND Suinesiaputra, Avan AND Lamata, Pablo AND Masci, Pier-Giorgio AND Young, Alistair A.},
title = { { Cardiac Shape-Derived Endophenotypes and Covariant Correction using Statistical Motion Atlases } },
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}
}