Fluoroscopic 3D Image Generation from Patient-Specific PCA Motion Models Derived from 4D-CBCT Patient Datasets: A...

Area of Science:

• Medical Imaging
• Radiotherapy Physics
• Computational Anatomy

Background:

• Four-dimensional cone-beam CT (4D-CBCT) provides time-varying volumetric data crucial for radiotherapy.
• Accurate patient anatomy and motion representation are vital for effective treatment delivery.
• Existing methods may not fully capture complex 3D non-rigid patient motion during treatment.

Purpose of the Study:

• To develop a method for generating fluoroscopic (time-varying) volumetric images using patient-specific motion models.
• To utilize four-dimensional cone-beam CT (4D-CBCT) data for deriving these motion models.
• To improve the accuracy of tumor and anatomical structure localization on the day of treatment.

Main Methods:

• Derivation of patient-specific motion models from 4D-CBCT image sets using deformable image registration (DIR).
• Application of Principal Component Analysis (PCA) to reduce dimensionality of displacement vector fields (DVFs) from DIR.
• Iterative optimization of PCA motion models by comparing real and simulated CBCT projections to generate fluoroscopic 3D images.

Main Results:

• The developed method successfully generated fluoroscopic 3D images accounting for patient motion.
• Tumor localization accuracy demonstrated a mean absolute error (MAE) of 2.29 mm (Patient 1) and 1.89 mm (Patient 2) in the SI direction.
• The 95th percentile error was 5.79 mm (Patient 1) and 4.82 mm (Patient 2), indicating good precision.

Conclusions:

• The study demonstrates the feasibility of 4D-CBCT-based PCA motion models for accounting for 3D non-rigid patient motion.
• This approach has the potential to accurately localize tumors and anatomical structures on the day of treatment.
• The method offers a promising tool for enhancing precision in image-guided radiotherapy.