Author: Robert A. Corns, Mohammad Kanber 👨🔬
Affiliation: East Carolina University, East Carolina University Brody School of Medicine 🌍
Purpose:
Accurate determination of the radiation isocenter is crucial for precise radiotherapy treatments, directly impacting patient safety and treatment quality. This study presents a computational framework to identify the radiation isocenter using multi-axis geometric data from film-based dosimetric measurements.
Methods:
Two films were taped to a 6 cm thick solid water phantom, aligned at the corner to maintain identical x, y coordinates, and placed on the LINAC table with the isocenter between the films. Thin radiation slits were created using the MLC, with each exposure producing an entrance on one film and an exit on the other across various gantry, collimator, and table angles. The films were scanned on a flatbed scanner for accurate registration.
Using ImageJ software, two points were digitized along each slit, providing four points per slit pair. An optimization-based plane-fitting method slightly adjusted the input points within predefined tolerances to minimize the sum of squared distances from the points to the fitted plane. Each axis of rotation had several planes associated with it. From these, three planes were selected to form a triangular tunnel, and spheres tangent to the planes were computed at each end by minimizing the maximum distance of the candidate sphere’s center to the planes. The axis was then estimated as the line joining the centers of the two spheres. Finally, a similar optimization approach identified a candidate sphere that touched all three axes, with its diameter scored for QA purposes. An SRS-ready LINAC should have a diameter less than 1 mm.
Results:
The optimization-based method demonstrated sub-millimeter accuracy, confirmed through 3D visualizations, with sphere diameters within the 1 mm threshold.
Conclusion:
This study introduces a robust framework for radiation isocenter determination, combining film-based data acquisition and optimization-based geometric modeling. The method enhances equipment calibration accuracy, improving radiotherapy treatment delivery and patient outcomes.