EUVIMEDAvasta EUVIMED
Open access

Dynamic cone beam CT reconstruction via spatiotemporal Gaussian neural representation

Yabo Fu, Hao Zhang, Weixing Cai, Huiqiao Xie, Licheng Kuo, Laura Cervino, Jean Moran, Xiang Li, Tianfang Li

Medical Physics · 2025

DOI: 10.1002/mp.70127 ↗

At a glance

Journal article in Medical Physics, 2025. Written by Yabo Fu and 8 co-authors. Institution: Department of Medical Physics, Memorial Sloan Kettering Cancer Center, New York, New York, USA. Major MEDLINE topics: Cone-Beam Computed Tomography, Four-Dimensional Computed Tomography, Image Processing, Computer-Assisted, Neural Networks, Computer. 29 references. The full text is freely available in PubMed Central. Funded by: NCI NIH HHS.

AbstraktSubjectsBibliograafilised andmedCiteRelated studies

Abstract (original language)

Background

In image‐guided radiotherapy (IGRT), four‐dimensional cone‐beam computed tomography (4D‐CBCT) is critical for assessing tumor motion during a patient's breathing cycle prior to beam delivery. However, generating 4D‐CBCT images with sufficient quality requires significantly more projection images than a standard 3D‐CBCT scan, leading to extended scanning times and increased imaging dose to the patient.

Purpose

To introduce a novel spatiotemporal Gaussian neural representation framework to reconstruct high‐temporal dynamic CBCT images from 1‐minute acquisition, preserving motion dynamics and fine spatial details without relying on prior images or motion models.

Methods

Our framework employs a differentiable 4D Gaussian representation initialized from average CBCT images. Gaussian points are characterized by position, covariance, rotation, and density, offering a compact and dynamic model for CBCT scenes. A Gaussian deformation network, incorporating a HexPlane encoder and multi‐head decoder, predicts Gaussian deformations to minimize L1 and structural similarity index measure (SSIM) losses between rendered and measured projections. Adaptive Gaussian control refines the representation by pruning underutilized Gaussians and densifying points in high‐gradient regions. The method was benchmarked on the AAPM SPARE challenge datasets and further validated with clinical CBCT scans from a Varian TrueBeam system. For the AAPM SPARE challenge datasets, the performance of the proposed method was evaluated using the root‐mean‐squared‐error (RMSE) and the structural similarity index (SSIM) in the four regions of interest: Body, Lung, PTV, and Bone. The geometric accuracy was evaluated by calculating the registration error when aligning the tumor to the ground truth using the Elastix package, focusing on pixels within the planning target volume (PTV). To demonstrate our method's capability in high‐temporal motion dynamic modeling using extremely undersampled projections, the clinical half‐fan projections from a 1‐minute Varian TrueBeam acquisition were sorted into 50 phases with approximately 18 projections per phase, significantly finer than the commonly used 10‐phase binning.

Results

Compared to the AAPM SPARE challenge participant methods, our method achieved superior geometric accuracy in terms of PTV alignment error, and comparable RMSE and SSIM when no prior 4DCT or motion model is used for our reconstruction. For PTV alignment, our method achieved translational and rotational errors of 0.54 mm (LR), 0.76 mm (SI), 1.36 mm (AP), 0.55° (rAP), and 0.93° (rSI), and 1.31° (rLR), respectively. For high temporal dynamic CBCT reconstruction, our method successfully reconstructed a 50‐phase CBCT from a 1‐minute Varian Truebeam half‐fan scan, demonstrating effective streak artifact suppression, respiratory motion preservation, and fine detail restoration. Reconstruction on a single NVIDIA RTX A6000 GPU required approximately 30–80 min, depending on the number of Gaussian points used (ranging from 50 to 400K), to reconstruct CBCT from 680 projections acquired with a 30 × 40 cm detector. Our code and reconstruction results can be found at: https://github.com/fuyabo/4DGS_for_4DCBCT/tree/main .

Conclusions

The spatiotemporal Gaussian framework is a novel data‐driven dynamic CBCT reconstruction technique that features excellent geometric accuracy in terms of PTV alignment and high‐temporal motion modeling, indicating promise for tumor motion assessment and high‐temporal respiratory motion modeling based on a 1‐minute half‐fan scan prior to beam delivery.

Subjects

MeSH terms (MEDLINE)

Cone-Beam Computed TomographyFour-Dimensional Computed TomographyImage Processing, Computer-AssistedNeural Networks, ComputerNormal DistributionHumansMovement

Author keywords

4DCBCT reconstructionspatiotemporal Gaussian representation

Authors and institutions

  • Department of Medical Physics, Memorial Sloan Kettering Cancer Center, New York, New York, USA.

Funding

  • NCI NIH HHS

Cite

APA

Fu, Y., Zhang, H., Cai, W., Xie, H., Kuo, L., Cervino, L., Moran, J., Li, X., & Li, T. (2025). Dynamic cone beam CT reconstruction via spatiotemporal Gaussian neural representation. Medical Physics. https://doi.org/10.1002/mp.70127

Vancouver

Fu Y, Zhang H, Cai W, Xie H, Kuo L, Cervino L, et al. Dynamic cone beam CT reconstruction via spatiotemporal Gaussian neural representation. Medical Physics. 2025. doi:10.1002/mp.70127

BibTeX
@article{Fu2025,
  author = {Fu, Yabo and Zhang, Hao and Cai, Weixing and Xie, Huiqiao and Kuo, Licheng and Cervino, Laura and Moran, Jean and Li, Xiang and Li, Tianfang},
  title = {Dynamic cone beam CT reconstruction via spatiotemporal Gaussian neural representation},
  journal = {Medical Physics},
  year = {2025},
  doi = {10.1002/mp.70127}
}
Start for free →