Tracking tumor boundary using point correspondence for adaptive radio therapy

Area of Science:

• Medical imaging
• Radiation oncology
• Computational anatomy

Background:

• Accurate tumor tracking in image-guided radiation therapy (IGRT) is crucial to prevent irradiating healthy tissues and ensure complete tumor coverage.
• Challenges in MRI-based tumor delineation include similar tissue appearances and organ motion, particularly respiratory movement.
• Precise tumor localization is vital for effective cancer treatment and minimizing side effects.

Purpose of the Study:

• To develop an automated approach for tracking the center and boundary of mobile tumor regions in MRI sequences.
• To address the challenges of delineating tumors in moving organs for adaptive radiation therapy (ART).
• To improve the accuracy of radiation delivery by accurately contouring the region of interest.

Main Methods:

• Utilized a nonrigid registration method and the RealTITracker algorithm for MRI tumor region tracking.
• Investigated two segmentation approaches: manual segmentation of the first frame versus all frames during pretreatment.
• Evaluated performance using a sequence of 600 MRI images from 6 patients.

Main Results:

• The method using manual segmentation of all frames combined with moving mesh-based registration achieved the best performance.
• Achieved an average Dice Score of 0.89 ± 0.04 and a Hausdorff Distance of 3.38 ± 0.10 mm.
• Demonstrated superior accuracy in tracking tumor regions affected by respiratory motion.

Conclusions:

• The developed boundary tracking tool shows promise for delineating tumor regions in the presence of respiratory movement.
• The approach is suitable for the constraints of adaptive radiation therapy.
• This tool can enhance the precision and effectiveness of IGRT.