Evaluation of automated image registration algorithm for image-guided radiotherapy (IGRT).

Area of Science:

• Medical Physics
• Radiotherapy Technology
• Image Analysis

Background:

• Accurate patient positioning is critical for effective image-guided radiation therapy (IGRT).
• Image registration (IR) software automates the detection of setup errors, enhancing treatment precision.
• Evaluating the performance of IR software under controlled conditions is essential for clinical implementation.

Purpose of the Study:

• To assess the accuracy and reproducibility of an image registration software for detecting simulated patient setup errors.
• To quantify the residual errors in translation and rotation after automatic image registration.
• To determine the overall performance of an image-guided radiation therapy (IGRT) system.

Main Methods:

• Simulated 27 setup errors (translations and rotations) on an ExactCouch phantom.
• Acquired orthogonal kV radiographs and cone-beam CT (CBCT) in half-fan and full-fan modes.
• Utilized 2D-2D and 3D-3D matching software based on mutual information for image registration.

Main Results:

• The automatic IR demonstrated excellent reproducibility with a mean standard deviation of 0.1 mm in translation and 0.0° in rotation.
• Average residual errors were within ±0.5 mm and ±0.2° for translational and rotational errors, respectively.
• The IGRT system achieved an accuracy within 1.8 mm and 0.4°, with high reproducibility under controlled conditions.

Conclusions:

• The evaluated image registration software is accurate and reproducible for detecting simulated setup errors in IGRT.
• The findings support the use of this IGRT system, with inherent process errors needing consideration in clinical protocols.
• This study provides valuable data for optimizing clinical IGRT protocols and ensuring patient safety.