Deformable Scintillation Dosimeter I: Challenges and Implementation using Computer Vision Techniques

TL;DR

This paper develops a computer vision-based method to correct signal variations in deformable scintillation dosimeters, enabling accurate real-time dose measurements in radiotherapy with complex geometries.

Contribution

It introduces a novel correction approach using stereo vision techniques for deformable scintillation dosimeters, improving accuracy and enabling real-time applications.

Findings

Signal variations can be effectively corrected using stereo vision.

The system achieves 0.008 cm position and 2° angle accuracy.

Corrections reduce intensity variation from ±10% to ±1%.

Abstract

Plastic scintillation detectors are increasingly used to measure dose distributions in the context of radiotherapy treatments. Their water-equivalence, real-time response and high spatial resolution distinguish them from traditional detectors, especially in complex irradiation geometries. Their range of applications could be further extended by embedding scintillators in a deformable matrix mimicking anatomical changes. In this work, we characterized signal variations arising from the translation and rotation of scintillating fibers with respect to a camera. Corrections are proposed using stereo vision techniques and two sCMOS complementing a CCD camera. The study was extended to the case of a prototype real-time deformable dosimeter comprising an array of 19 scintillating. The signal to angle relationship follows a gaussian distribution (FWHM = 52{\deg}) whereas the intensity variation from radial displacement follows the inverse square law. Tracking the position and angle of the fibers enabled the correction of these spatial dependencies. The detecting system provides an accuracy and precision of respectively 0.008 cm and 0.03 cm on the position detection. This resulted in an uncertainty of 2{\deg} on the angle measurement. Displacing the dosimeter by $\pm$3 cm in depth resulted in relative intensities of 100$\pm$10% (mean $\pm$ standard deviation) to the reference position. Applying corrections reduced the variations thus resulting in relative intensities of 100$\pm$1%. Similarly, for lateral displacements of $\pm$3 cm, intensities went from 98$\pm$3% to 100$\pm$1% after the correction. Therefore, accurate correction of the signal collected by a camera imaging the output of scintillating elements in a 3D volume is possible. This work paves the way to the development of real-time scintillator-based deformable dosimeters.