Learning Dense 2D-3D Correspondence for X-ray-to-CT Registration of Knee Bones

• URL: http://arxiv.org/abs/2607.22803v1
• Date: Fri, 24 Jul 2026 15:24:05 GMT
• Title: Learning Dense 2D-3D Correspondence for X-ray-to-CT Registration of Knee Bones
• Abstract summary: Recovering the 6-DoF pose of the knee bones from a radiograph turns a low-dose image into a quantitative measurement of joint geometry.<n>We learn an amortized, subject-agnostic dense 2D-3D correspondence, supervised solely by projection geometry.
• Score: 1.2974485228990196
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Recovering the 6-DoF pose of the knee bones from a plain radiograph, given the patient's segmented pre-operative CT, turns a routine low-dose image into a quantitative measurement of joint geometry, without the added dose of a repeat CT or a fixed biplanar rig. Classic solutions align a rendered bone silhouette to image edges; recent alternatives refine pose by backpropagating an image-similarity loss through a differentiable X-ray renderer. Both operate one patient at a time and are fragile under a single view. Silhouettes are depth-ambiguous, and differentiable-rendering refinement has a narrow capture range at substantial per-iteration cost. We instead learn an amortized, subject-agnostic dense 2D-3D correspondence, supervised solely by projection geometry. One shared-weight model per bone, trained across 758 patients, registers patients unseen during training. The pose then follows in closed form from a global, initialization-free, render-free PnP+RANSAC solve. Because X-ray formation is transmissive, our correspondence target is transmission-aware rather than tied to a single surface. Though trained only to register, the representation is anatomically semantic: a simple classifier reads a landmark's anatomical region from its embedding across held-out patients, and the same features separate the knee's bones into a 2D-3D-consistent identity learned without any bone label. On a large single-institution cohort the model generalizes well to held-out patients.

Related papers

• Rendering Novel Views of MRI Using 3D Gaussian Splatting [54.6286495315908]
  We adapt 3D Gaussian Splatting to form a reconstruction starting from sparse anisotropic MRIs, and imaging planes aligned with the anatomy relevant for clinical evaluation are then sampled and rendered.<n>The novel view plane is optimal for diagnostic radiological grading of the target anatomy, whereas the original MRI is not.
  arXiv  Detail & Related papers  (2026-06-24T18:00:03Z)
• Toward Training-Free Zero-Shot Anomaly Detection in 3D Medical Images: A Batch-Based Approach Using 2D Foundation Models [0.0]
  We propose CS3F, a training-free framework for anomaly detection in 3D medical images using 2D foundation models.<n>Anomaly scores are obtained from cross-subject mutual similarity.<n>CS3F is evaluated on brain MRI across metastases, glioma, and stroke, as well as validated on lung CT to test generalizability.
  arXiv  Detail & Related papers  (2026-06-17T06:52:07Z)
• VRXU-net: A Deep Learning Approach for Brain Ischemic Stroke Lesion Detection and Segmentation in T1W MRI [0.0]
  This study introduces a novel VRU-Net architecture, derived from visual features, residual connections, and a U-shaped network, for detecting and segmenting ischemic stroke lesions in 3D magnetic resonance imaging scans.<n>To improve both performance and processing speed, a high-performance classifier is applied before the segmentation model in a sequential framework.<n>The proposed model is trained on the Anatomical Tracings of Lesions After Stroke dataset and outperforms state-of-the-art models in terms of accuracy and Dice coefficient.
  arXiv  Detail & Related papers  (2026-05-20T18:44:08Z)
• HyDAR-Pano3D: A Hybrid Disentangled Anatomical Recovery Framework for Panoramic-to-3D Reconstruction [10.481519508246437]
  We propose HyDAR-Pano3D, a two-stage framework that reformulates PR-to-CBCT reconstruction as a disentangled anatomical recovery problem.<n>In Stage 1, a dual-encoder network integrates radiographic features with SAM-derived semantic priors to reconstruct an arch-normalized canonical volume.<n>In Stage 2, an Anatomical Restoration Network predicts a prior-constrained structured deformation field to map this canonical volume back to the native space.
  arXiv  Detail & Related papers  (2026-05-20T07:20:56Z)
• Preoperative-to-intraoperative Liver Registration for Laparoscopic Surgery via Latent-Grounded Correspondence Constraints [51.7011449975586]
  Land-Reg is a deformable registration framework that learns latent-grounded 2D-3D landmark correspondences.<n>For rigid registration, Land-Reg embraces a Cross-modal Latent Alignment module.<n>An Uncertainty-enhanced Overlap Landmark Detector with similarity matching is proposed to robustly estimate explicit 2D-3D landmark correspondences.
  arXiv  Detail & Related papers  (2026-03-02T10:44:03Z)
• On the Localization of Ultrasound Image Slices within Point Distribution Models [84.27083443424408]
  Thyroid disorders are most commonly diagnosed using high-resolution Ultrasound (US) Longitudinal tracking is a pivotal diagnostic protocol for monitoring changes in pathological thyroid morphology. We present a framework for automated US image slice localization within a 3D shape representation.
  arXiv  Detail & Related papers  (2023-09-01T10:10:46Z)
• XraySyn: Realistic View Synthesis From a Single Radiograph Through CT Priors [118.27130593216096]
  A radiograph visualizes the internal anatomy of a patient through the use of X-ray, which projects 3D information onto a 2D plane. To the best of our knowledge, this is the first work on radiograph view synthesis. We show that by gaining an understanding of radiography in 3D space, our method can be applied to radiograph bone extraction and suppression without groundtruth bone labels.
  arXiv  Detail & Related papers  (2020-12-04T05:08:53Z)
• Three-dimensional Segmentation of the Scoliotic Spine from MRI using Unsupervised Volume-based MR-CT Synthesis [3.6273410177512275]
  We present an unsupervised, fully three-dimensional (3D) cross-modality synthesis method for segmenting scoliotic spines. A 3D CycleGAN model is trained for an unpaired volume-to-volume translation across MR and CT domains. The resulting segmentation is used to reconstruct a 3D model of the spine.
  arXiv  Detail & Related papers  (2020-11-25T18:34:52Z)
• Convolutional-LSTM for Multi-Image to Single Output Medical Prediction [55.41644538483948]
  A common scenario in developing countries is to have the volume metadata lost due multiple reasons. It is possible to get a multi-image to single diagnostic model which mimics human doctor diagnostic process.
  arXiv  Detail & Related papers  (2020-10-20T04:30:09Z)
• MotioNet: 3D Human Motion Reconstruction from Monocular Video with Skeleton Consistency [72.82534577726334]
  We introduce MotioNet, a deep neural network that directly reconstructs the motion of a 3D human skeleton from monocular video. Our method is the first data-driven approach that directly outputs a kinematic skeleton, which is a complete, commonly used, motion representation.
  arXiv  Detail & Related papers  (2020-06-22T08:50:09Z)
• End-To-End Convolutional Neural Network for 3D Reconstruction of Knee Bones From Bi-Planar X-Ray Images [6.645111950779666]
  We present an end-to-end Convolutional Neural Network (CNN) approach for 3D reconstruction of knee bones directly from two bi-planar X-ray images.
  arXiv  Detail & Related papers  (2020-04-02T08:37:11Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.