Imaging-anchored Multiomics in Cardiovascular Disease: Integrating Cardiac Imaging, Bulk, Single-cell, and Spatial Transcriptomics

TL;DR

The paper addresses the challenge that cardiovascular imaging and molecular profiling are often analyzed separately, hindering mechanistic understanding and translation. It advocates an imaging-anchored framework that uses cardiac imaging as the spatial reference and spatial transcriptomics as the bridge to link voxel- and region-level imaging features with cell-type programs and molecular states. It surveys representation-learning approaches across imaging, bulk and single-cell omics, and spatial transcriptomics, along with fusion strategies (early/intermediate/late fusion, graph-based methods, contrastive learning, and multimodal transformers), to construct joint latent spaces and integrative pipelines for radiogenomics, spatial alignment, and virtual transcriptomics. The review highlights representative frameworks (e.g., MOFA+, scVI, DCCA, MVAE, PoE-VAEs), publicly available datasets, and practical considerations such as data harmonisation, reproducibility and hardware costs, while outlining open challenges and future directions toward spatial generative models, foundation models and digital-twin–driven cardiovascular translation. Overall, it provides a roadmap for moving from descriptive imaging toward mechanistic, molecularly informed cardiovascular analysis with potential for patient stratification and targeted therapies. The work emphasizes interdisciplinary collaboration among cardiology, computational biology and AI to realize imaging-anchored multiomics at scale.

Abstract

Cardiovascular disease arises from interactions between inherited risk, molecular programmes, and tissue-scale remodelling that are observed clinically through imaging. Health systems now routinely generate large volumes of cardiac MRI, CT and echocardiography together with bulk, single-cell and spatial transcriptomics, yet these data are still analysed in separate pipelines. This review examines joint representations that link cardiac imaging phenotypes to transcriptomic and spatially resolved molecular states. An imaging-anchored perspective is adopted in which echocardiography, cardiac MRI and CT define a spatial phenotype of the heart, and bulk, single-cell and spatial transcriptomics provide cell-type- and location-specific molecular context. The biological and technical characteristics of these modalities are first summarised, and representation-learning strategies for each are outlined. Multimodal fusion approaches are reviewed, with emphasis on handling missing data, limited sample size, and batch effects. Finally, integrative pipelines for radiogenomics, spatial molecular alignment, and image-based prediction of gene expression are discussed, together with common failure modes, practical considerations, and open challenges. Spatial multiomics of human myocardium and atherosclerotic plaque, single-cell and spatial foundation models, and multimodal medical foundation models are collectively bringing imaging-anchored multiomics closer to large-scale cardiovascular translation.

Figures (4)

• Figure 1: Imaging-anchored multiomics framework. Cardiac imaging modalities (echocardiography, cardiac MRI and cardiac CT) define a spatial phenotype of the heart that can be linked to bulk, single-cell and spatial transcriptomic measurements. Joint representations support downstream tasks such as multimodal fusion, radiogenomics, spatial alignment and image-based prediction of gene-expression patterns.
• Figure 2: Multimodal fusion architectures for imaging–omics integration. Panel A: early fusion concatenates imaging and omics feature vectors followed by a single predictive model. Panel B: intermediate fusion uses modality-specific encoders and a shared latent space, often with reconstruction heads. Panel C: late fusion combines predictions from separate imaging and omics models via an ensemble or calibration layer. Panel D: graph-based and contrastive approaches construct cross-modal latent spaces using biological or spatial graphs and contrastive objectives; multimodal transformers can be viewed as token-based generalisations of these designs.
• Figure 3: Archetypal pipelines for imaging–omics integration. Radiogenomics workflows link imaging-derived features to bulk or single-cell molecular profiles via feature extraction, joint latent modelling and association or causal inference; spatial molecular alignment pipelines register in vivo imaging, ex vivo imaging, histology and spatial transcriptomics into a common coordinate system; virtual transcriptomics approaches use deep models to predict molecular modules or spatial expression maps directly from imaging inputs.
• Figure 4: Translational roadmap from imaging-anchored multiomics to digital twins. Top lane: model stack comprising modality-specific encoders, multimodal fusion models and transformer or foundation models architectures. Middle lane: an agentic layer in which large language models orchestrate tool calls to encoders, fusion models and knowledge bases to generate structured reports and explanations. Bottom lane: translational path from data curation and model development through external validation and deployment to decision support and digital twin--enabled in silico trials.