AnyCXR: Human Anatomy Segmentation of Chest X-ray at Any Acquisition Position using Multi-stage Domain Randomized Synthetic Data with Imperfect Annotations and Conditional Joint Annotation Regularization Learning

TL;DR

AnyCXR tackles the lack of generalizable chest X-ray segmentation across arbitrary projection angles by pairing a large-scale synthetic data generator (MSDR) with a partial-label–aware learning framework (CAR). MSDR creates over 100k diverse, anatomically aligned DRRs from 3D CTs, accompanied by a two-stage quality-control pipeline to curate reliable annotations. CAR leverages partial supervision by enforcing latent-space anatomical priors and reconstruction consistency, enabling a unified segmentation of 54 structures across PA, LA, and oblique views. The approach yields strong zero-shot performance on real CXRs and demonstrates tangible clinical utility in automated CTR, spine curvature assessment, and disease classification, suggesting AnyCXR can serve as a scalable anatomy-aware foundation for CXR AI systems.

Abstract

Robust anatomical segmentation of chest X-rays (CXRs) remains challenging due to the scarcity of comprehensive annotations and the substantial variability of real-world acquisition conditions. We propose AnyCXR, a unified framework that enables generalizable multi-organ segmentation across arbitrary CXR projection angles using only synthetic supervision. The method combines a Multi-stage Domain Randomization (MSDR) engine, which generates over 100,000 anatomically faithful and highly diverse synthetic radiographs from 3D CT volumes, with a Conditional Joint Annotation Regularization (CAR) learning strategy that leverages partial and imperfect labels by enforcing anatomical consistency in a latent space. Trained entirely on synthetic data, AnyCXR achieves strong zero-shot generalization on multiple real-world datasets, providing accurate delineation of 54 anatomical structures in PA, lateral, and oblique views. The resulting segmentation maps support downstream clinical tasks, including automated cardiothoracic ratio estimation, spine curvature assessment, and disease classification, where the incorporation of anatomical priors improves diagnostic performance. These results demonstrate that AnyCXR establishes a scalable and reliable foundation for anatomy-aware CXR analysis and offers a practical pathway toward reducing annotation burdens while improving robustness across diverse imaging conditions.

Figures (12)

• Figure 1: AnyCXR Data Pipeline: This workflow projects 3D CT volumes into diverse 2D Digitally Reconstructed Radiographs (DRRs) and segmentation masks using Multi-stage Domain Randomization. The pipeline includes an automated filtering step that discards invalid masks (red). The feature plot (bottom left) demonstrates that this method generates a data distribution significantly broader than typical real-world datasets.
• Figure 2: MSDR Pipeline: This flow chart details the Multi-stage Domain Randomization engine, which converts 3D CT data into 2D synthetic DRRs through a sequence of stochastic steps. The pipeline involves Pre-projection Randomizations acting on Hounsfield Units and image artifacts, In-projection Randomizations modifying geometry, and Post-projection Randomizations simulating detector response and inverting polarity. The red boxes denote randomization names, while blue boxes denote different approaches.
• Figure 3: Distribution of organ annotations after filtering: This chart shows the anatomical label presence percentage across the 2,982 curated patient CT volumes after the two-stage quality control process. The distribution is grouped by major anatomical categories (Bone, Heart and Vessels, Lung, Rib, and Vertebrae), demonstrating the partial yet high-quality nature of the training dataset used for AnyCXR.
• Figure 4: DRR Augmentation Demonstration for Major Innovation: This figure visually illustrates the effect of the primary Multi-stage randomization on the resulting DRR, starting from a Default DRR (left panel). Each subsequent column shows a different major 3D randomization applied in isolation: Bone-to-soft-tissue, Inter-bone, Intra-bone, Soft-tissue, and Projection geometry randomization, demonstrating how these 3D perturbations introduce diverse image appearances and contrast variations.
• Figure 5: CAR Framework Overview: Input X-rays are segmented by a U-Net, with predictions substituting missing ground truth labels. Conditioned on image $\mathbf{X}$, these maps are encoded by $G_{\theta_1}$ to align their latent distributions via $\mathcal{L}_{\text{Dist}}$ (optimizing the U-Net), while the decoder $G_{\theta_2}$ ensures latent integrity via the reconstruction loss $\mathcal{L}_{\text{Recon}}$.