VoxelSynth3D reduces metal artifacts in musculoskeletal CT scans
VoxelSynth3D: Interpretable Volumetric Image-Domain Metal Artifact Reduction with a Paired Synthetic CLINIC-Metal Benchmark
Computer Vision and Pattern Recognition
Summary
Metal objects inside the body, like implants, can cause blurry or distorted areas in CT images, making it hard to see bones and soft tissues clearly. The authors propose VoxelSynth3D, a method that improves CT images by fixing these blurry areas without needing extra data or training. They also created a synthetic dataset to test how well different methods work at removing these metal artifacts. Their approach gently improves image quality near the metal while keeping important implant details intact.
What this means in practice
- •For medical imaging software developers: Integrate VoxelSynth3D into CT software to improve bone and soft tissue visibility around implants without raw projection data.
- •For clinical radiology departments: Use the synthetic CLINIC-Metal dataset to validate and benchmark metal artifact reduction tools in musculoskeletal CT imaging.
Authors
Amritesh Banerjee, Abdul Basit, Renil Renji Joseph, Nouhaila Innan, Muhammad Shafique
Abstract
Metal artifacts in postoperative musculoskeletal CT obscure bone-implant and adjacent soft-tissue interfaces. Many metal artifact reduction (MAR) methods require unavailable raw projections or learned models that may shift across scanners and implants. We present VoxelSynth3D, a training-free 3D image-domain framework for reconstructed CT. The framework combines support masking, normalized tissue synthesis, deviation gating, and restricted edge refinement. Detected implant voxels are preserved in the output, while correction targets metal-induced artifacts in the surrounding tissue. We also construct Synthetic CLINIC-Metal, a controlled paired synthetic evaluation resource, from no-metal CTPelvic1K volumes with clean targets, metal/artifact masks, fixed seeds, and patient-level splits; 75 unpaired real metal cases receive qualitative/no-reference evaluation only. The operating point was fixed in a near-flat validation basin. With exact-mask oracle localization, all methods share a metal-excluded tissue ROI. On 40 held-out cases, VoxelSynth3D reduced RMSE from 801.48 to 786.18 HU (paired gain 15.30 HU, 95% CI 11.68-19.23), improving every case and exceeding the evaluated 3D Gaussian smoother by 13.58 HU. Clean-edge agreement decreased next to metal but exceeded input beyond 5 mm. Thus, VoxelSynth3D provides case-consistent within-distribution tissue-error reduction with a localized structural tradeoff. Spacing-aware sensitivity retained aggregate broad-region improvement and identified near-metal calibration as a target.