Directional sinogram interpolation for motion weighted 4D cone-beam CT reconstruction

Area of Science:

• Medical Imaging
• Radiotherapy Physics
• Image Reconstruction

Background:

• Four-dimensional (4D) cone-beam CT (CBCT) is essential for image-guided radiation therapy.
• Image quality in 4D CBCT is often degraded by streak artifacts due to insufficient projections and patient motion.
• Existing reconstruction methods struggle to adequately address motion-induced artifacts and preserve image quality.

Purpose of the Study:

• To propose and evaluate a novel motion weighted reconstruction (MWR) method for improving 4D CBCT image quality.
• To reduce streak artifacts and image blur in 4D CBCT.
• To enhance the accuracy of radiation therapy planning by providing clearer images of moving targets.

Main Methods:

• Developed a motion weighted reconstruction (MWR) method involving projection interpolation (directional sinogram interpolation - DSI) and deformable image registration for motion estimation.
• Reconstructed a baseline 3D CBCT using the FDK algorithm from non-interpolated projections.
• Combined the interpolated 4D CBCT and the 3D FDK CBCT using voxel-based weights derived from local motion estimation.
• Compared MWR with conventional 4D CBCT and McKinnon and Bates (MKB) reconstruction methods using phantom and patient data.

Main Results:

• MWR reduced root-mean-square-error (RMSE) by 38.7% in a 4D phantom, compared to 24.7% for MKB, relative to conventional 4D CBCT.
• Image blur in static regions was minimal with MWR, and less pronounced in moving regions compared to MKB and 3D FDK methods.
• Average contrast-to-noise ratio (CNR) in lung cancer patients improved by a factor of 3.5 with MWR, compared to 1.7 for MKB and 2.8 for DSI, relative to 4D FDK.

Conclusions:

• The proposed motion weighted reconstruction (MWR) method effectively reduces artifacts and improves image quality in 4D CBCT.
• MWR enhances image clarity in both static and respiratory-moving regions, outperforming conventional 4D CBCT and MKB methods.
• This technique holds significant potential for improving the precision and efficacy of image-guided radiation therapy.