CT to cone-beam CT deformable registration with simultaneous intensity correction.

Area of Science:

• Medical Imaging
• Radiotherapy Physics
• Computational Imaging

Background:

• Deformable image registration (DIR) is essential for adaptive radiation therapy.
• Intensity-based registration methods like demons struggle with CT-CBCT DIR due to intensity inconsistencies.
• CBCT images often contain artifacts that further complicate registration.

Purpose of the Study:

• To develop a novel algorithm, deformation with intensity simultaneously corrected (DISC), to address CT-CBCT DIR challenges.
• To improve the accuracy and robustness of deformable image registration in adaptive radiation therapy.
• To ensure intensity consistency between CT and CBCT images during registration.

Main Methods:

• Proposed a variant of the demons algorithm, DISC, incorporating adaptive intensity correction.
• Implemented an intensity correction step that matches voxel intensity moments within local patches between CT and CBCT.
• Utilized a compute unified device architecture (CUDA) for efficient GPU implementation.

Main Results:

• DISC demonstrated robustness against CBCT artifacts and intensity variations.
• The algorithm significantly improved registration accuracy compared to the standard demons algorithm.
• Evaluated on simulated data and six clinical head-and-neck cancer patient cases.

Conclusions:

• DISC effectively handles intensity inconsistencies and artifacts in CT-CBCT DIR.
• The proposed method enhances the accuracy of deformable image registration for adaptive radiation therapy.
• DISC offers a promising solution for improving treatment precision in radiotherapy.