| Question 1: Compared with the conventional manual QA processes, which of the following descriptions about autonomous QA processes is correct: |
| Reference: | Gilmer Valdes, Olivier Morin, Yanisley Valenciaga, Niel Kirby, Jean Pouliot, Cynthia Chuang: Use of TrueBeam developer mode for imaging QA, Journal of Applied Clinical Medical Physics, 16 (4), 322-333(2015) |
| Choice A: | More efficient, but less accurate |
| Choice B: | More efficient, but complicated and unstable |
| Choice C: | More efficient, more stable and accurate |
| Choice D: | Less efficient, but more stable and accurate |
| Question 2: Typically, an autonomous QA process includes: |
| Reference: | Cesare H Jenkins, Dominik J Naczynski, Shu-Jung S Yu, Yong Yang and Lei Xing, Automating quality assurance of digital linear accelerators using a radioluminescent phosphor coated phantom and optical imaging, Phys. Med. Biol. 61, L29-L37 (2016) |
| Choice A: | Automatic data acquisition |
| Choice B: | Automatic data processing |
| Choice C: | Automatic analysis and reporting |
| Choice D: | All of the above |
| Question 3: Which statement is true about why machine learning methods are useful in physics QA? |
| Reference: | Baozhou Sun, Dao Lam, Deshan Yang, Kevin Grantham, Tiezhi Zhang, Sasa Mutic, Tianyu Zhao, A machine learning approach to the accurate prediction of monitor units for a compact proton machine, Medical Physics, March 2018 |
| Choice A: | To predict physics QA passing rates, machine learning methods could be more accurate than the conventional methods (i.e. multi-variable linear fitting). |
| Choice B: | For detecting errors in the patient treatment plan parameters, the machine learning models can detect more errors than conventional rule-based error detection methods. |
| Choice C: | Both A and B are true. |
| Question 4: Select the best machine learning method for extracting knowledge (for example, lung + IMRT -> prescription = 60 Gy) from patient datasets. |
| Reference: | Altaf et al., Applications of association rule mining in health informatics: a survey, Journal Artificial Intelligence Review, vol 47 Issue 3, March 2017 |
| Choice A: | Bayesian network |
| Choice B: | Association rules |
| Choice C: | Decision trees |
| Choice D: | Artificial neural networks |
| Question 5: Which QA data storage method is best suited for future data analysis, mining, and sharing? |
| Reference: | M.Y.Y. Law, B. Liu, L.W. Chan, Informatics in Radiology—DICOM-RT based electronic patient record information system for radiation therapy, Radiographics, 29 (2009), pp. 961–972 |
| Choice A: | Portable document format (PDF) |
| Choice B: | Database |
| Choice C: | Excel spreadsheets |
| Choice D: | Paper |
| Question 6: What benefit does QA protocol standardization provide? |
| Reference: | Santanam, L. et al. Standardizing naming conventions in radiation oncology. Int. J. Radiat. Oncol. Biol. Phys. 83, 1344–1349 (2012) |
| Choice A: | Improved communication between multiple clinics |
| Choice B: | Ensures that efficient QA procedures are being used |
| Choice C: | Improves data analysis |
| Choice D: | All of the above |
