A minimax entropy registration framework for patient setup verification in radiotherapy

Area of Science:

• Medical Physics
• Radiotherapy Imaging
• Image Registration

Background:

• Accurate patient setup is crucial in external beam radiotherapy (EBRT) for precise dose delivery.
• Fractionated treatments require consistent verification over time.
• Portal images and CT datasets are commonly used for setup verification.

Purpose of the Study:

• To develop and present a minimax entropy registration framework for patient setup verification in EBRT.
• To enhance the accuracy and robustness of portal image registration to 3D CT data.
• To address challenges posed by low-resolution and low-contrast portal images.

Main Methods:

• Development of a minimax entropy registration framework.
• Utilizing dual portal images and 3D CT datasets for registration.
• Employing iteratively and automatically estimated segmentation of portal images.
• Guiding registration using bony structures in portal and CT images.
• Proposing ultrasound imaging to quantify prostate motion relative to pelvic bone.

Main Results:

• An overview of the registration framework is presented.
• Initial testing of the proposed approach is described.
• The framework aims for more accurate and robust registration despite portal image limitations.
• Ultrasound integration is proposed to account for prostate movement.

Conclusions:

• The developed minimax entropy registration framework offers a novel approach to patient setup verification in EBRT.
• The method leverages bony anatomy for registration but acknowledges limitations due to soft tissue movement.
• Incorporating ultrasound imaging is proposed to improve accuracy for treatments involving mobile organs like the prostate.