From Baseline to Follow-Up: Counterfactual Spine DXA Image Synthesis in UK Biobank Using a Causal Hierarchical Variational Autoencoder

2026-05-21

AI Summary

This work addresses the challenge of generating anatomically plausible and controllable counterfactual images to characterize the influence of specific factors—such as age—on spinal morphology in large-scale skeletal imaging. The authors propose the Causal Hierarchical Variational Autoencoder (CHVAE), which integrates participant metadata from the UK Biobank with lumbar spine dual-energy X-ray absorptiometry (DXA) images. By leveraging the abduction–action–prediction framework from causal inference, CHVAE enables controlled synthesis of vertebral shape changes under age-based interventions. Experimental results demonstrate that the counterfactual images produced by the model exhibit strong agreement with actual longitudinal measurements across key anatomical metrics, substantially enhancing both anatomical plausibility and interpretability of the generated outcomes.

Abstract

Dual-energy X-ray absorptiometry (DXA) is widely used for large-scale skeletal assessment, yet learning controllable and interpretable factor-specific anatomical variation remains challenging. We propose a metadata-conditioned causal hierarchical variational autoencoder (CHVAE) for causally consistent generation of anteroposterior (AP) spine DXA images from the UK Biobank (UKB). The model is trained on 3,743 raw AP spine scans from the first imaging visit and conditioned on basic participant attributes and lumbar morphometry. Causal consistency is evaluated in a baseline-to-follow-up setting using abduction--action--prediction (AAP): latent variables are abducted from baseline images, age is intervened to the repeat-imaging value, and the resulting counterfactual follow-up morphometry is compared with observed repeat-imaging measurements. Results show strong absolute-level agreement for key vertebral morphometry variables under age intervention, supporting intervention-aligned synthesis of anatomically plausible DXA images.

Problem

• counterfactual image synthesis
• spine DXA
• causal consistency
• anatomical variation
• baseline-to-follow-up

Innovation

• causal hierarchical variational autoencoder
• counterfactual image synthesis
• DXA spine imaging
• abduction-action-prediction
• morphometric consistency