High-Resolution Cranial Defect Reconstruction by Iterative, Low-Resolution, Point Cloud Completion Transformers

TL;DR

This paper introduces an iterative transformer-based point cloud completion method for high-resolution cranial defect reconstruction, offering a faster, resource-efficient alternative to volumetric approaches with superior performance.

Contribution

The work reformulates cranial defect reconstruction as a point cloud completion task and proposes a novel iterative transformer method that handles high-resolution data efficiently.

Findings

Outperforms volumetric methods in GPU memory efficiency

Maintains high-quality defect reconstruction at any resolution

Faster training and inference compared to existing approaches

Abstract

Each year thousands of people suffer from various types of cranial injuries and require personalized implants whose manual design is expensive and time-consuming. Therefore, an automatic, dedicated system to increase the availability of personalized cranial reconstruction is highly desirable. The problem of the automatic cranial defect reconstruction can be formulated as the shape completion task and solved using dedicated deep networks. Currently, the most common approach is to use the volumetric representation and apply deep networks dedicated to image segmentation. However, this approach has several limitations and does not scale well into high-resolution volumes, nor takes into account the data sparsity. In our work, we reformulate the problem into a point cloud completion task. We propose an iterative, transformer-based method to reconstruct the cranial defect at any resolution while also being fast and resource-efficient during training and inference. We compare the proposed methods to the state-of-the-art volumetric approaches and show superior performance in terms of GPU memory consumption while maintaining high-quality of the reconstructed defects.