Intraoperative 2D/3D Image Registration via Differentiable X-ray Rendering
- URL: http://arxiv.org/abs/2312.06358v2
- Date: Wed, 27 Mar 2024 12:24:29 GMT
- Title: Intraoperative 2D/3D Image Registration via Differentiable X-ray Rendering
- Authors: Vivek Gopalakrishnan, Neel Dey, Polina Golland,
- Abstract summary: We present DiffPose, a self-supervised approach that leverages patient-specific simulation and differentiable physics-based rendering to achieve accurate 2D/3D registration without relying on manually labeled data.
DiffPose achieves sub-millimeter accuracy across surgical datasets at intraoperative speeds, improving upon existing unsupervised methods by an order of magnitude and even outperforming supervised baselines.
- Score: 5.617649111108429
- License: http://creativecommons.org/licenses/by/4.0/
- Abstract: Surgical decisions are informed by aligning rapid portable 2D intraoperative images (e.g., X-rays) to a high-fidelity 3D preoperative reference scan (e.g., CT). 2D/3D image registration often fails in practice: conventional optimization methods are prohibitively slow and susceptible to local minima, while neural networks trained on small datasets fail on new patients or require impractical landmark supervision. We present DiffPose, a self-supervised approach that leverages patient-specific simulation and differentiable physics-based rendering to achieve accurate 2D/3D registration without relying on manually labeled data. Preoperatively, a CNN is trained to regress the pose of a randomly oriented synthetic X-ray rendered from the preoperative CT. The CNN then initializes rapid intraoperative test-time optimization that uses the differentiable X-ray renderer to refine the solution. Our work further proposes several geometrically principled methods for sampling camera poses from $\mathbf{SE}(3)$, for sparse differentiable rendering, and for driving registration in the tangent space $\mathfrak{se}(3)$ with geodesic and multiscale locality-sensitive losses. DiffPose achieves sub-millimeter accuracy across surgical datasets at intraoperative speeds, improving upon existing unsupervised methods by an order of magnitude and even outperforming supervised baselines. Our code is available at https://github.com/eigenvivek/DiffPose.
Related papers
- Intraoperative Registration by Cross-Modal Inverse Neural Rendering [61.687068931599846]
We present a novel approach for 3D/2D intraoperative registration during neurosurgery via cross-modal inverse neural rendering.
Our approach separates implicit neural representation into two components, handling anatomical structure preoperatively and appearance intraoperatively.
We tested our method on retrospective patients' data from clinical cases, showing that our method outperforms state-of-the-art while meeting current clinical standards for registration.
arXiv Detail & Related papers (2024-09-18T13:40:59Z) - R$^2$-Gaussian: Rectifying Radiative Gaussian Splatting for Tomographic Reconstruction [53.19869886963333]
3D Gaussian splatting (3DGS) has shown promising results in rendering image and surface reconstruction.
This paper introduces R2$-Gaussian, the first 3DGS-based framework for sparse-view tomographic reconstruction.
arXiv Detail & Related papers (2024-05-31T08:39:02Z) - X-Ray to CT Rigid Registration Using Scene Coordinate Regression [1.1687067206676627]
This paper proposes a fully automatic registration method that is robust to extreme viewpoints.
It is based on a fully convolutional neural network (CNN) that regresses the overlapping coordinates for a given X-ray image.
The proposed method achieved an average mean target registration error (mTRE) of 3.79 mm in the 50th percentile of the simulated test dataset and projected mTRE of 9.65 mm in the 50th percentile of real fluoroscopic images for pelvis registration.
arXiv Detail & Related papers (2023-11-25T17:48:46Z) - On the Localization of Ultrasound Image Slices within Point Distribution
Models [84.27083443424408]
Thyroid disorders are most commonly diagnosed using high-resolution Ultrasound (US)
Longitudinal tracking is a pivotal diagnostic protocol for monitoring changes in pathological thyroid morphology.
We present a framework for automated US image slice localization within a 3D shape representation.
arXiv Detail & Related papers (2023-09-01T10:10:46Z) - The Impact of Loss Functions and Scene Representations for 3D/2D
Registration on Single-view Fluoroscopic X-ray Pose Estimation [1.758213853394712]
We first develop a differentiable projection rendering framework for the efficient computation of Digitally Reconstructed Radiographs (DRRs)
We then perform pose estimation by iterative descent using various candidate loss functions, that quantify the image discrepancy of the synthesized DRR with respect to the ground-truth fluoroscopic X-ray image.
Using the Mutual Information loss, a comprehensive evaluation of pose estimation performed on a tomographic X-ray dataset of 50 patients$'$ skulls shows that utilizing either discretized (CBCT) or neural (NeTT/mNeRF) scene representations in DiffProj leads to
arXiv Detail & Related papers (2023-08-01T01:12:29Z) - Geometry-Aware Attenuation Learning for Sparse-View CBCT Reconstruction [53.93674177236367]
Cone Beam Computed Tomography (CBCT) plays a vital role in clinical imaging.
Traditional methods typically require hundreds of 2D X-ray projections to reconstruct a high-quality 3D CBCT image.
This has led to a growing interest in sparse-view CBCT reconstruction to reduce radiation doses.
We introduce a novel geometry-aware encoder-decoder framework to solve this problem.
arXiv Detail & Related papers (2023-03-26T14:38:42Z) - CNN-based real-time 2D-3D deformable registration from a single X-ray
projection [2.1198879079315573]
This paper presents a method for real-time 2D-3D non-rigid registration using a single fluoroscopic image.
A dataset composed of displacement fields and 2D projections of the anatomy is generated from a preoperative scan.
A neural network is trained to recover the unknown 3D displacement field from a single projection image.
arXiv Detail & Related papers (2022-12-15T09:57:19Z) - Transfer Learning from an Artificial Radiograph-landmark Dataset for
Registration of the Anatomic Skull Model to Dual Fluoroscopic X-ray Images [0.4205692673448206]
We propose a transfer learning strategy for 3D-to-2D registration using deep neural networks trained from an artificial dataset.
Digitally reconstructed radiographs (DRRs) and radiographic skull landmarks were automatically created from craniocervical CT data of a female subject.
They were used to train a residual network (ResNet) for landmark detection and a cycle generative adversarial network (GAN) to eliminate the style difference between DRRs and actual X-rays.
The methodology to strategically augment artificial training data can tackle the complicated skull registration scenario, and has potentials to extend to widespread registration scenarios.
arXiv Detail & Related papers (2021-08-14T04:49:36Z) - Deep Learning compatible Differentiable X-ray Projections for Inverse
Rendering [8.926091372824942]
We propose a differentiable by deriving the distance travelled by a ray inside mesh structures to generate a distance map.
We show its application by solving the inverse problem, namely reconstructing 3D models from real 2D fluoroscopy images of the pelvis.
arXiv Detail & Related papers (2021-02-04T22:06:05Z) - Revisiting 3D Context Modeling with Supervised Pre-training for
Universal Lesion Detection in CT Slices [48.85784310158493]
We propose a Modified Pseudo-3D Feature Pyramid Network (MP3D FPN) to efficiently extract 3D context enhanced 2D features for universal lesion detection in CT slices.
With the novel pre-training method, the proposed MP3D FPN achieves state-of-the-art detection performance on the DeepLesion dataset.
The proposed 3D pre-trained weights can potentially be used to boost the performance of other 3D medical image analysis tasks.
arXiv Detail & Related papers (2020-12-16T07:11:16Z) - Neural Descent for Visual 3D Human Pose and Shape [67.01050349629053]
We present deep neural network methodology to reconstruct the 3d pose and shape of people, given an input RGB image.
We rely on a recently introduced, expressivefull body statistical 3d human model, GHUM, trained end-to-end.
Central to our methodology, is a learning to learn and optimize approach, referred to as HUmanNeural Descent (HUND), which avoids both second-order differentiation.
arXiv Detail & Related papers (2020-08-16T13:38:41Z)
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.