WE-E-213CD-07: Deformable Registration Between CT and Truncated CBCT for Adaptive Therapy Dose Calculation

Area of Science:

• Medical Physics
• Image Processing
• Radiation Oncology

Background:

• Deformable image registration (DIR) is essential for adaptive radiation therapy (ART), enabling dose calculation and contour propagation by aligning planning CT with current CBCT.
• Truncated CBCT images, due to limited field of view (FOV), pose challenges for DIR, often yielding unphysical results and potential dose calculation errors.

Purpose of the Study:

• To develop and evaluate a novel method to enhance existing DIR algorithms, specifically addressing the problem of image truncation in CBCT for ART.
• To improve the accuracy of DIR when registering planning CT with truncated CBCT images.

Main Methods:

• The method estimates the CBCT's FOV radius and employs the Deformation with Intensity Simultaneously Corrected (DISC) algorithm for deformation field calculation.
• During DISC iterations, deformation vector fields outside the FOV are smoothed and propagated from within the FOV.
• The approach was evaluated using six head-and-neck and two prostate cancer cases, with subsequent dose calculations to assess impact.

Main Results:

• The enhanced DIR method showed improved accuracy, with average increases in NMI, NCC, and FSIM compared to standard DISC.
• Significant reduction in the relative L2 dose distance was observed, decreasing from 9.25% to 1.41% within the FOV.
• These results indicate a substantial improvement in dose distribution accuracy.

Conclusions:

• A deformation field propagation method was successfully developed to improve DIR for registering planning CT with small FOV CBCT images.
• The algorithm demonstrated enhanced registration accuracy for dose calculation in head-and-neck and prostate cancer cases.
• The findings suggest this method is valuable for improving ART outcomes by addressing CBCT truncation issues.