Deep Learning-based Point Cloud Registration for Augmented Reality-guided Surgery

TL;DR

This paper evaluates deep learning methods for point cloud registration in augmented reality-guided surgery, finding that traditional methods still outperform deep learning on medical datasets.

Contribution

It introduces a dataset of medical and AR point clouds and compares deep learning models with conventional registration techniques in this context.

Findings

Deep learning methods show potential but are not yet superior.

Conventional registration pipelines outperform deep learning models on the dataset.

The dataset bridges medical imaging and AR point cloud data.

Abstract

Point cloud registration aligns 3D point clouds using spatial transformations. It is an important task in computer vision, with applications in areas such as augmented reality (AR) and medical imaging. This work explores the intersection of two research trends: the integration of AR into image-guided surgery and the use of deep learning for point cloud registration. The main objective is to evaluate the feasibility of applying deep learning-based point cloud registration methods for image-to-patient registration in augmented reality-guided surgery. We created a dataset of point clouds from medical imaging and corresponding point clouds captured with a popular AR device, the HoloLens 2. We evaluate three well-established deep learning models in registering these data pairs. While we find that some deep learning methods show promise, we show that a conventional registration pipeline still outperforms them on our challenging dataset.