Premium
SU‐G‐201‐03: Automation of High Dose Rate Brachytherapy Quality Assurance: Development of a Radioluminescent Detection System for Simultaneous Detection of Activity, Timing, and Positioning
Author(s) -
Jenkins C,
Xing L,
Fahimian B
Publication year - 2016
Publication title -
medical physics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.473
H-Index - 180
eISSN - 2473-4209
pISSN - 0094-2405
DOI - 10.1118/1.4956876
Subject(s) - quality assurance , brachytherapy , calibration , timestamp , computer science , dwell time , pixel , automation , intensity (physics) , computer vision , materials science , artificial intelligence , real time computing , optics , physics , engineering , medicine , mechanical engineering , clinical psychology , operations management , external quality assessment , quantum mechanics , radiation therapy
Purpose: Accuracy of positioning, timing and activity is of critical importance for High Dose Rate (HDR) brachytherapy delivery. Respective measurements via film autoradiography, stop‐watches and well chambers can be cumbersome, crude or lack dynamic source evaluation capabilities. To address such limitations, a single device radioluminescent detection system enabling automated real‐time quantification of activity, position and timing accuracy is presented and experimentally evaluated. Methods: A radioluminescent sheet was fabricated by mixing Gd?O?S:Tb with PDMS and incorporated into a 3D printed device where it was fixated below a CMOS digital camera. An Ir‐192 HDR source (VS2000, VariSource iX) with an effective active length of 5 mm was introduced using a 17‐gauge stainless steel needle below the sheet. Pixel intensity values for determining activity were taken from an ROI centered on the source location. A calibration curve relating intensity values to activity was generated and used to evaluate automated activity determination with data gathered over 6 weeks. Positioning measurements were performed by integrating images for an entire delivery and fitting peaks to the resulting profile. Timing measurements were performed by evaluating source location and timestamps from individual images. Results: Average predicted activity error over 6 weeks was .35 ± .5%. The distance between four dwell positions was determined by the automated system to be 1.99 ± .02 cm. The result from autoradiography was 2.00 ± .03 cm. The system achieved a time resolution of 10 msec and determined the dwell time to be 1.01 sec ± .02 sec. Conclusion: The system was able to successfully perform automated detection of activity, positioning and timing concurrently under a single setup. Relative to radiochromic and radiographic film‐based autoradiography, which can only provide a static evaluation positioning, optical detection of temporary radiation induced luminescence enables dynamic detection of position enabling automated quantification of timing with millisecond accuracy.