Cone-beam CT reconstruction for non-periodic organ motion using time-ordered chain graph model

Area of Science:

• Medical Imaging
• Radiology
• Computational Imaging

Background:

• Four-dimensional cone-beam computed tomography (4D CBCT) is crucial for visualizing organ motion during radiation therapy.
• Existing reconstruction methods like total variation-based compressed sensing (TVCS) and prior-image constrained compressed sensing (PICCS) have limitations in accurately capturing non-periodic motion.

Purpose of the Study:

• To introduce and evaluate a novel four-dimensional (4D) cone-beam computed tomography (CBCT) reconstruction method using the time-ordered chain graph model (TCGM).
• To compare the performance of the TCGM approach against TVCS and PICCS for non-periodic organ motion.
• To assess the clinical feasibility of TCGM for reconstructing time-series images of organ motion.

Main Methods:

• Iterative reconstruction using the TCGM was performed with 90° projection ranges.
• The TCGM method was compared with TVCS (200° projection ranges) and PICCS (90° projection ranges).
• Reconstruction performance was evaluated using an elliptic-cylindrical digital phantom with simulated motion and clinical patient data from the pelvic region.

Main Results:

• The TCGM demonstrated superior temporal resolution, evidenced by narrower penumbral widths in phantom studies compared to TVCS and PICCS.
• TCGM achieved 18.2% and 23.1% narrower penumbral widths than TVCS for 3 cm/30 s and 3 cm/60 s motion, respectively.
• Clinical data showed that TCGM provided visibly clearer images with reduced blurring for motion related to rectal gas and stool.

Conclusions:

• The TCGM is a feasible and effective method for visualizing non-periodic organ motion in 4D CBCT.
• The TCGM significantly improves temporal resolution for 4D image series reconstruction compared to TVCS and PICCS.
• The TCGM facilitates clear visualization of intrafraction motion, such as rectal gas and flatus, in clinical patient data.