Development of a Methodology for Clinical in vivo Optical Dosimetry and Uncertainty Analysis
Abstract
Purpose
Cherenkov imaging is a technique used to observe surface dose deposition during radiation therapy. While proportional to dose, the imaged signal cannot derive absolute dose, due to the attenuation and inhomogeneity of human tissue optics. However, controlled, scintillating discs can be placed on the patient’s surface and used to quantify delivered dose. While this has been an area of research for several years, a standardized method has yet to be established along with its associated uncertainties. The purpose of this project was to establish a workflow and identify the uncertainties along with associated correction factors.
Methods
The proposed workflow is focused on placing the scintillating discs in regions of ambiguous dose initially highlighted by Cherenkov imaging, i.e. slight Cherenkov light detected on a patient’s contralateral breast. Specialized software implemented with Cherenkov cameras can image the scintillation signal and convert it to absolute dose through a calibration factor generated by repeated irradiation of the disc to known doses to form a calibration curve. Initial uncertainties were assessed through repeated 200MU at 600MU/min measurements of a single scintillator along with comparing scintillators from the same batch. Additionally, camera optics/electronics lead to spatial inhomogeneity within the scene. A diffuse light panel was imaged to characterize this effect and generate a correction method.
Results
Initial uncertainties were processed to demonstrate the capabilities of the approach. Repeated single scintillator measurements produced a k=1 uncertainty of 0.39%. 2.8% uncertainty was found between scintillators within a batch and dropped by 0.86% when accounting for radiological thickness of individual scintillators. Spatial inhomogeneity uncertainty dropped from 15.6% to 2.3% after correction.
Conclusion
The potential for real-time dosimetry using Cherenkov imaging cameras already installed at many institutions is an exciting development exhibiting reasonable uncertainties. Work is ongoing to identify every source of uncertainty and bring them within acceptable limits.