A technique for estimating 4D-CBCT using prior knowledge and limited-angle projections.

Area of Science:

• Medical Physics
• Radiotherapy Technology
• Image Reconstruction

Background:

• Accurate 4D-CBCT (4-dimensional computed tomography) is crucial for lung radiotherapy target verification.
• Estimating onboard 4D-CBCT from limited-angle projections using prior information remains a challenge.

Purpose of the Study:

• To develop and evaluate a technique for estimating onboard 4D-CBCT using prior information and limited-angle projections.
• To enable potential 4D target verification during lung radiotherapy.

Main Methods:

• A motion modeling and free-form deformation (MM-FD) technique was developed, utilizing principal component analysis (PCA) for motion model extraction.
• Deformation field maps (DFMs) were optimized by matching digitally reconstructed radiographs to limited-angle projections.
• The 4D digital extended-cardiac-torso phantom and simulated lung patient data were used for evaluation, comparing MM-FD against MM-only and FD-only techniques.

Main Results:

• The MM-FD technique significantly improved estimation accuracy, reducing mean volume percentage-difference (VPD) and center-of-mass shift (COMS) compared to MM-only and FD-only methods.
• With orthogonal-view 30° scan angles, MM-FD achieved a mean VPD of 5.22% and COMS of 0.5 mm.
• Accuracy was affected by projection angle, sampling interval, and noise level, with MM-FD demonstrating robustness under various simulated conditions.

Conclusions:

• The MM-FD technique substantially enhances onboard 4D-CBCT estimation accuracy using prior 4D-CT and limited-angle projections.
• This technique shows promise for inter- and intrafractional 4D target localization verification in lung radiotherapy.