Learning-Based Patch-Wise Metal Segmentation with Consistency Check

TL;DR

This paper introduces a learning-based patch-wise metal segmentation method with a novel consistency check that improves segmentation accuracy and reduces false positives in 3D X-ray metal artifact reduction.

Contribution

It proposes a new patch-wise segmentation network combined with a consistency check as a post-processing step, enhancing segmentation performance and reliability.

Findings

Achieved an average IoU score of 0.924 on test data.

Significantly reduces false positive segmentations.

Ensures consistent segmentation results.

Abstract

Metal implants that are inserted into the patient's body during trauma interventions cause heavy artifacts in 3D X-ray acquisitions. Metal Artifact Reduction (MAR) methods, whose first step is always a segmentation of the present metal objects, try to remove these artifacts. Thereby, the segmentation is a crucial task which has strong influence on the MAR's outcome. This study proposes and evaluates a learning-based patch-wise segmentation network and a newly proposed Consistency Check as post-processing step. The combination of the learned segmentation and Consistency Check reaches a high segmentation performance with an average IoU score of 0.924 on the test set. Furthermore, the Consistency Check proves the ability to significantly reduce false positive segmentations whilst simultaneously ensuring consistent segmentations.