Motion compensated cone-beam CT reconstruction using ana priorimotion model from CT simulation: a pilot study.

Area of Science:

• Medical Physics
• Radiotherapy Technology
• Image Reconstruction

Background:

• Traditional 4D-CBCT reconstruction methods struggle with motion artifacts.
• Existing motion-compensated SART (MC-SART) requires sufficient projections, which are often lacking.
• A novel approach is needed to improve CBCT image quality for accurate tumor targeting.

Purpose of the Study:

• To demonstrate the feasibility of integrating a CT simulation-derived motion model into MC-SART for improved CBCT reconstruction.
• To introduce and evaluate a model-based CBCT (MB-CBCT) technique for enhanced motion compensation.
• To assess the impact of MB-CBCT on image sharpness and tumor delineation accuracy.

Main Methods:

• Acquired 5D CT datasets and pre-treatment CBCT with simultaneous breathing surrogates from 5 patients.
• Cross-calibrated 5D CT and CBCT breathing waveforms using diaphragm matching.
• Employed 5D CT motion model parameters within MC-SART, introducing Amplitude Reassignment Motion Modeling.
• Evaluated tumor and diaphragm sharpness, comparing MB-CBCT against 4D-CBCT and MC-SART.

Main Results:

• Diaphragm sharpness consistently improved with increasing amplitude resolution in 4/5 patients.
• MB-CBCT demonstrated consistently high image quality across multiple treatment fractions.
• Tumor centroids showed stable locations with an average difference of 0.74 ± 0.31 mm between 4D-CBCT and MB-CBCT.
• MB-CBCT significantly improved diaphragm sharpness and overall image quality compared to 3D-CBCT and 4D-CBCT.

Conclusions:

• Prior 5D CT motion models can effectively compensate for motion during CBCT reconstruction.
• MB-CBCT represents a significant advancement over traditional MC-SART and 4D-CBCT in image quality.
• This technique shows promise for improving the accuracy and reliability of image-guided radiation therapy.