H3D-MarNet: Wavelet-Guided Dual-Path Learning for Metal Artifact Suppression and CT Modality Transformation for Radiotherapy Workflows

2026-05-12

AI Summary

Metal artifacts severely degrade CT image quality, compromising radiotherapy planning and diagnostic accuracy. This work proposes a two-stage framework to address this challenge: first, wavelet-based frequency-domain preprocessing suppresses artifacts, followed by a dual-path network that integrates CNNs and Transformers to perform cross-modality synthesis from kVCT to MVCT. The architecture employs attention mechanisms to fuse multi-scale features and incorporates deep-supervised, multi-stage decoding to reconstruct high-fidelity images. By innovatively combining wavelet denoising with a hybrid CNN–Transformer design, the method effectively suppresses artifacts while preserving anatomical structures, achieving a PSNR of 28.14 dB and an SSIM of 0.717 on artifact-contaminated slices—significantly enhancing both image quality and clinical utility.

Abstract

Metal artifacts in computed tomography (CT) severely degrade image quality, compromising diagnostic accuracy and radiotherapy planning, especially in cancer patients with high-density implants. We propose H3D-MarNet, a two-stage framework for artifact-aware CT domain transformation from kilo-voltage CT (kVCT) to mega-voltage CT (MVCT). In the first stage, a wavelet-based preprocessing module suppresses metal-induced artifacts through frequency-aware denoising while preserving anatomical structures. In second stage, Domain-TransNet performs kVCT-to-MVCT domain transformation using a hybrid volumetric learning architecture. Domain-TransNet integrates a CNN-based encoder to capture fine-grained local anatomical details and a transformer-based encoder to model long-range volumetric dependencies. The complementary representations are fused through an attention-based feature fusion mechanism to ensure spatial and contextual coherence across slices. A multi-stage, attention-guided decoder, supported by deep supervision, progressively reconstructs artifact-suppressed MVCT volumes. Extensive experiments demonstrate that H3D-MarNet achieves 28.14 dB PSNR and 0.717 SSIM on artifact-affected slices from full dataset, indicating effective metal artifact suppression and anatomical preservation, highlighting its potential for reliable CT modality transformation in clinical radiotherapy workflows.

Problem

• metal artifact
• CT modality transformation
• radiotherapy
• image quality degradation
• kVCT-to-MVCT

Innovation

• wavelet-based denoising
• dual-path learning
• kVCT-to-MVCT transformation
• CNN-transformer fusion
• attention-guided decoding