Organ-contour-driven auto-matching algorithm in image-guided radiotherapy

Area of Science:

• Medical Physics
• Radiation Oncology
• Image Processing

Background:

• Image-guided radiotherapy (IGRT) requires precise patient positioning.
• Organ contour information can enhance the accuracy of IGRT positioning.
• Cervical cancer treatment involves complex target volumes and organs at risk.

Purpose of the Study:

• To evaluate the clinical applicability of an organ-contour-driven auto-matching algorithm for IGRT.
• To assess the algorithm's ability to generate consistent and accurate patient positioning.
• To determine the impact of rotational corrections on planning target volume (PTV) coverage.

Main Methods:

• Developed a weight-based auto-matching algorithm using mesh models of daily and reference MRI scans.
• Implemented a cost function prioritizing specific organs for positioning in three scenarios.
• Optimized translation and rotation values for cervix and uterus, with a rotation limit of ±3°.
• Evaluated PTV coverage using isotropic 5 mm and anisotropic margins.

Main Results:

• Translations exceeding 10 mm occurred in 9-18% of fractions, varying by scenario.
• Over 80% of fractions achieved ≥80% coverage for small PTVs, and >90% achieved ≥95% coverage for large PTVs.
• Rotational corrections (±3°) had a minimal impact (<1% difference in median coverage) compared to translational positioning alone.

Conclusions:

• The organ-contour-driven algorithm enables quantitative and prioritized organ positioning in IGRT.
• Consistent displacement values were algorithmically generated, demonstrating reliability.
• Limited rotational corrections (±3°) minimally affected PTV coverage, simplifying the positioning process.