Learning-Based Patch-Wise Metal Segmentation with Consistency Check
This addresses metal artifact reduction in medical imaging, which is crucial for improving diagnostic accuracy in trauma interventions, but it is incremental as it builds on existing segmentation methods.
The study tackled metal artifact reduction in 3D X-ray images by developing a learning-based patch-wise segmentation network with a Consistency Check post-processing step, achieving an average IoU score of 0.924 on the test set and reducing false positives.
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.