Program Information
Machine Learning for Automated Quality Assurance in Radiotherapy
I El Naqa1*, J DeMarco2 , H Al-Hallaq3 , J Booth4 , T Ritter5 , G Kim6 , S Park7 , R Popple8 , M Perez9 , K Farrey3 , J Moran1 , (1) University of Michigan, Ann Arbor, MI, (2) Cedars-Sinai Medical Center, Los Angeles, CA, (3) The University of Chicago, Chicago, IL, (4) Royal North Shore Hospital, St Leonards, ,(5) VCU Health System, Ann Arbor, MI, (6) University of California, San Diego, La Jolla, CA, (7) McLaren-Flint, Flint, MI, (8) Univ Alabama Birmingham, Birmingham, AL, (9) Royal North Shore Hospital, St Leonards, New South Wales
Presentations
TU-FG-605-6 (Tuesday, August 1, 2017) 1:45 PM - 3:45 PM Room: 605
Purpose: Developing automated methods to identify task-driven quality assurance (QA) procedures is key towards increasing safety and efficiency. We investigate the use of machine learning methods for automation/targeting of QA and assessing its performance in multi-institutional data.
Methods: To enable automated analysis of QA data given its higher dimensional nature, we used nonlinear kernel mapping with support vector machines (SVM). Instead of using labeled data, which requires exhaustive annotation, we applied a clustering extension of SVM. QA test data are mapped by a Gaussian kernel to higher-dimensional feature space and searched for the minimal enclosing sphere. This sphere, when mapped back to the input data space along the principal components, can separate the data into several components, each enclosing a separate cluster of QA points that could be used to evaluate tolerance boundaries and test reliability. We evaluated this approach for gantry sag and radiation field shift measurements using EPID.
Results: Data from 8 LINACS and 7 institutions (n=119) were collected. A standardized EPID image of a phantom with fiducials provided deviation estimates between the radiation field and phantom center at 4 cardinal gantry angles. Deviation measurements in the horizontal direction (0°, 180°) were used to determine the gantry sag and deviations in the vertical direction (90°, 270°) were used to determine the field shift. These measurements were fed into the SVM clustering with varying sphere radii. For gantry sag analysis, a sphere radius=0.5 yielded two clusters with one identifying acceptable measurements and the other identifying outliers (2.5%), which contrasts with TG-142 limits of 1 mm. In the case of field shifts, SVM clustering identified two distinct classes of measurements primarily driven by variations in the second principal component at 270°.
Conclusion: Machine learning methods are promising for developing automated QA tools and providing insights into their reliability and reproducibility.
Funding Support, Disclosures, and Conflict of Interest: This work was partly supported by Varian Medical Systems.
Contact Email: