Point-Wise Geometry-Aware Transformer for Partial-to-Full Point Cloud Registration in Computer-Assisted Surgery

2026-06-11

AI Summary

This work addresses the challenges of partial-to-complete point cloud registration in surgical settings, where varying overlap ratios, non-uniform point density, and noise significantly degrade performance. To overcome these issues, the authors propose GAPR-Net, a coarse-to-fine architecture that innovatively embeds transformation-invariant per-point geometric features into a Transformer framework. By integrating convolutional operations with cross-attention mechanisms, the method effectively fuses local geometric details with global contextual information. Evaluated on four skeletal datasets, GAPR-Net demonstrates substantial improvements in robustness and accuracy, achieving a recall of 94.2%, an RMSE of 1.992 mm, and R² coefficients of 0.908 and 0.974 for rotation and translation prediction, respectively.

Abstract

Partial-to-full registration remains challenging due to varying overlap ratios, fluctuating point densities, and the presence of noise. While transformers have shown strong potential for point cloud processing, prior methods typically confine them to global context aggregation, overlooking fine-grained local geometry crucial for accurate correspondence. We propose \emph{GAPR-Net}, a learning-based point cloud registration framework with a coarse-to-fine architecture that combines convolution and transformer modules, in which local and global information is fused between the partial and full point clouds using a cross-attention mechanism. To achieve this, a transformation-invariant point-wise geometric feature representation is proposed, which can robustly capture relative geometric features for individual points with respect to their neighboring points. To evaluate the effectiveness of the proposed approach, experiments are conducted on four geometrically distinct bones, including the tibia, femur, pelvis, and thoracic cartilage. The overall registration recall reaches 94.2\%, the method results in a low RMSE of 1.992 mm and $R^2$ values of 0.908 and 0.974 for rotation and translation, respectively. The results demonstrate that the proposed method effectively addresses the partial-to-full point cloud registration problem. The proposed method enables highly accurate 3D point cloud registration using partial observation, providing a critical foundation for precise surgical navigation and robotic interventions in computer-assisted surgery. The code will be accessed after the double-blind review process.

Problem

• partial-to-full registration
• point cloud
• computer-assisted surgery
• geometric correspondence
• overlap variation

Innovation

• point-wise geometry
• cross-attention
• partial-to-full registration
• transformation-invariant features
• coarse-to-fine architecture