Deep Learning Enables Large-Scale Shape and Appearance Modeling in Total-Body DXA Imaging

2025-08-13

AI Summary

Manual annotation of anatomical landmarks in large-scale total-body dual-energy X-ray absorptiometry (TBDXA) images is time-consuming and exhibits poor inter-rater reliability. Method: We propose the first deep learning framework for fully automated keypoint localization across five distinct imaging modalities, enabling scalable body shape and appearance modeling (SAM). Trained on 1,683 high-quality manually annotated TBDXA scans, the model achieves 99.5% landmark localization accuracy on an external validation set. Contribution/Results: We construct a comprehensive SAM database comprising 35,928 TBDXA scans. Using Kolmogorov–Smirnov tests, we systematically demonstrate statistically significant associations between SAM-derived morphometric features and clinical biomarkers of frailty, metabolism, inflammation, and cardiometabolic health—thereby proposing a novel “body composition–morphology–health” hypothesis. All code, pretrained model weights, and modeling data are publicly released.

Technology Category

• Machine Learning: Multimodal Learning
• Constraint Satisfaction and Optimization: Satisfiability Modulo Theories
• Search and Optimization: Sampling/Simulation-based Search

Application Category

• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies
• Security and Privacy: Large-scale security measurements
• Economics, Online Markets and Human Computation: Data quality aspects of human-annotated datasets

Abstract

Total-body dual X-ray absorptiometry (TBDXA) imaging is a relatively low-cost whole-body imaging modality, widely used for body composition assessment. We develop and validate a deep learning method for automatic fiducial point placement on TBDXA scans using 1,683 manually-annotated TBDXA scans. The method achieves 99.5% percentage correct keypoints in an external testing dataset. To demonstrate the value for shape and appearance modeling (SAM), our method is used to place keypoints on 35,928 scans for five different TBDXA imaging modes, then associations with health markers are tested in two cohorts not used for SAM model generation using two-sample Kolmogorov-Smirnov tests. SAM feature distributions associated with health biomarkers are shown to corroborate existing evidence and generate new hypotheses on body composition and shape's relationship to various frailty, metabolic, inflammation, and cardiometabolic health markers. Evaluation scripts, model weights, automatic point file generation code, and triangulation files are available at https://github.com/hawaii-ai/dxa-pointplacement.

Problem

• Automates fiducial point placement on TBDXA scans
• Models body shape/appearance for health marker associations
• Validates method on large-scale datasets and cohorts

Innovation

• Deep learning for automatic fiducial point placement
• Validated on 1,683 manually-annotated TBDXA scans
• Applied to 35,928 scans for shape modeling