Dual-domain U-Nets with embedded back projection operators for motion-resolved 4D CBCT reconstruction

2026-08-04

AI Summary

This work addresses the limitations of conventional 4D CBCT reconstruction, which relies on respiratory signals and projection binning, leading to prolonged scan times, elevated radiation dose, and susceptibility to motion and sparse-sampling artifacts. The authors propose an end-to-end deep learning framework that directly reconstructs motion-resolved 4D CBCT from free-breathing scans without requiring respiratory signals or explicit binning. The core innovation is a dual-domain U-Net architecture: its encoder processes stacks of filtered projections, while the decoder operates in the volume domain, with multiscale, non-trainable backprojection operators embedded in the skip connections to enable joint feature fusion across projection and volume domains. Experiments demonstrate that the method achieves image quality on par with SART-TV on simulated data (reducing RMSE by 1.19 HU and improving PSNR by 0.09 dB), and in clinical assessments, 59% of experts preferred its tumor visualization and 47% favored its esophageal depiction, citing clearer dynamic structures and markedly reduced motion artifacts.

Abstract

Four-dimensional cone beam CT (4D CBCT) is important for image-guided radiation therapy of thoracic cancers, but its use is limited by long scan times, causing high patient dose and motion/sparse-sampling artifacts. We propose a deep learning method for motion-resolved 4D CBCT reconstruction from conventional free-breathing scans, without a respiratory signal or explicit projection binning. Our CNN takes free-breathing 3D CBCT projections as input and predicts a static volume at maximum inhalation plus ten displacement vector fields (DVFs) spanning a breathing cycle. The network extends U-Net: the encoder acts on filtered projection stacks, the decoder acts in the volume domain, and skip connections are replaced with non-trainable back-projection functions at multiple resolutions to transfer features between domains. The model is trained on simulated CBCT scans and evaluated on 11 unseen simulated patients and 13 clinical free-breathing scans. Two additional models (60 s and 6 s scans) were evaluated by clinical experts on three and two scans, comparing single phases of our 4D reconstruction to reference 3D SART-TV images for tumor and esophagus visibility. Experts preferred our method for tumor visibility (59% vs. 36% no preference, 5% reference) and esophagus visibility (47% vs. 42%, 11%). On simulated data, image quality matched SART-TV (mean RMSE: -1.19 HU, PSNR: +0.09 dB, SSIM: -0.009) while enabling 4D reconstruction. On clinical scans, our method showed sharper dynamic structures (e.g., diaphragm) and fewer motion streak artifacts than traditional reconstruction. This non-patient-specific CNN predicts static volumes and full 4D respiratory motion models from a single free-breathing scan, without a respiratory surrogate or projection binning, reducing motion artifacts while adding motion-modeling capability.

Problem

• 4D CBCT
• motion artifacts
• sparse-sampling artifacts
• image-guided radiation therapy
• thoracic cancers

Innovation

• dual-domain U-Net
• back-projection operator
• motion-resolved 4D CBCT
• displacement vector fields
• free-breathing reconstruction