Abstract
Multi-omics-driven health digital twins require interpretable representations that capture individual biological states across molecular and clinical dimensions. However, most multimodal models predict risk scores or biomarkers without explicitly linking molecular features to clinical endpoints, nor explaining how these relationships vary across individuals versus populations . In this preliminary study, we propose a population -to-personal network framework as an interpretable foundational layer fo r health digital twins. The framework maps longitudinal multi-omics and clinical features onto endpoint -centered networks anchored by clinically meaningful pivots. A population reference network is constructed from harmonized subject-timepoint snapshots to define cohort-level feature–endpoint relations. Personalized networks are then calibrated using individual longitudinal observations within the same feature-by-pivot structure. By contrasting these networks, the framework quantifies personal deviations an d extracts endpoint -specific deviation subnetwork s. A glucose -metabolism-focused case study demonstrates how individual-specific multi-omics associations can be visualized within a shared population coordinate system. Although this preliminary work does not infer causality or simulate interventions, it provides a preliminary and testable representation . Future work will test generalizability across prospective cohorts and evaluate clinical utility in real-world deployment.
Links to Paper and Supplementary Materials
Main Paper (Open Access Version): https://papers.miccai.org/miccai-2026-sat/paper/DT4H_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/forum?id=YSKrbidtxO
BibTex
@InProceedings{WeiZhe_PopulationtoPersonal_MICCAISAT2026,
author = { Wei, Zheng AND Liu, Jiaxin AND Wang, Avery Shuo AND Dong, Yiwei AND Deng, Wansi AND Zhang, Xueli AND Shang, Xianwen AND Zhou, Xin AND He, Mingguang AND Zhang, Jing},
title = { { Population-to-Personal Networks as an Interpretable Foundation for Multi-omics-Driven Health Digital Twins } },
booktitle = {Medical Image Computing and Computer Assisted Intervention -- MICCAI 2026 Workshops and Challenges},
year = {2026},
publisher = {Springer Nature Switzerland},
volume = {LNCS 17275},
month = {pending},
page = {pending}
}