Deep learning‐based dual‐domain reconstruction for nonstop gated CBCT in respiratory gating lung SBRT
Medical Physics · 2026
DOI: 10.1002/mp.70484 ↗Kurz gefasst
Fachartikel in der Zeitschrift Medical Physics, 2026. Verfasst von Mitchell Yu und 14 weiteren Autor:innen. Beteiligte Einrichtung: Department of Medical Physics, Memorial Sloan Kettering Cancer Center, New York, USA. Hauptthemen laut MEDLINE: Image Processing, Computer-Assisted, Cone-Beam Computed Tomography, Radiosurgery, Deep Learning. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 34 Literaturangaben. Gefördert von: NIH/NCI Cancer Center, Varian Medical Systems, AAPM Summer Undergraduate Fellowship.
Abstract (Originalsprache)
Background
The nonstop gated CBCT (ngCBCT) technique has been proposed as a next‐generation replacement to the current inefficient clinical gated CBCT (gCBCT), reducing half‐fan scan time from 2–8 min to 1 min on C‐arm linear accelerators while substantially lowering imaging dose. However, ngCBCT yields highly non‐uniform and under‐sampled projections, posing a major challenge for conventional reconstruction methods to achieve high‐quality images.
Purpose
To develop a powerful and efficient Dual‐Domain Convolutional Neural Network (DDCNN) tailored to the unique characteristics of ngCBCT projections to achieve high‐quality ngCBCT imaging.
Methods
Clinical raw gCBCT projections from 31 free‐breathing respiratory‐gated lung SBRT patients (77 half‐fan and 65 full‐fan scans) were retrospectively retrieved and down‐sampled based on respiratory signals to emulate ngCBCT acquisitions. The proposed DDCNN integrates a projection‐domain network, which completes missing projection data, with an image‐domain network that further reduces artifacts. The two domains are linked by the fast FDK algorithm, enabling accurate and efficient reconstruction suitable for real‐time clinical use.
Results
DDCNN outperformed conventional reconstruction methods and image‐domain‐only CNN approaches both qualitatively and quantitatively. It achieved high image quality and rapid reconstruction (
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Department of Medical Physics, Memorial Sloan Kettering Cancer Center, New York, USA.
- Department of Radiation Oncology, Memorial Sloan Kettering Cancer Center, New York, USA.
- Department of Radiation Oncology, Stanford University School of Medicine, Palo Alto, Stanford, California, USA.
Förderung
- NIH/NCI Cancer Center
- Varian Medical Systems
- AAPM Summer Undergraduate Fellowship
Zitieren
APA
Yu, M., Berry, S., Silverberg, N., Fu, Y., Harris, W., Cai, W., Kuo, L., He, X., Gelblum, D., Mueller, B., Cervino, L., Li, T., Li, X., Moran, J., & Zhang, H. (2026). Deep learning‐based dual‐domain reconstruction for nonstop gated CBCT in respiratory gating lung SBRT. Medical Physics. https://doi.org/10.1002/mp.70484
Vancouver
Yu M, Berry S, Silverberg N, Fu Y, Harris W, Cai W, et al. Deep learning‐based dual‐domain reconstruction for nonstop gated CBCT in respiratory gating lung SBRT. Medical Physics. 2026. doi:10.1002/mp.70484
BibTeX
@article{Yu2026,
author = {Yu, Mitchell and Berry, Sean and Silverberg, Noah and Fu, Yabo and Harris, Wendy and Cai, Weixing and Kuo, LiCheng and He, Xiuxiu and Gelblum, Daphna and Mueller, Boris and Cervino, Laura and Li, Tianfang and Li, Xiang and Moran, Jean and Zhang, Hao},
title = {Deep learning‐based dual‐domain reconstruction for nonstop gated CBCT in respiratory gating lung SBRT},
journal = {Medical Physics},
year = {2026},
doi = {10.1002/mp.70484}
}
EUVIMED