| Question 1: In knowledge-based dose-volume histogram (DVH) prediction, DVHs of OARs for a resultant plan can be correlated with specific: |
| Reference: | Reference: Ma M, et al: Dosimetric features‐driven machine learning model for DVH prediction in VMAT treatment planning. Medical Physics, 46(2), 857-867, 2019. |
| Choice A: | Geometric features |
| Choice B: | Anatomical features |
| Choice C: | Dosimetric features |
| Choice D: | All above |
| Question 2: Which of the following is true for script-based automatic plan check? A script-based automatic planning check tool can check: |
| Reference: | Reference: Liu S et al: Optimizing efficiency and safety in external beam radiotherapy using automated plan check (APC) tool and six sigma methodology. JCAMP. 20, 8: 56–64, 2019 |
| Choice A: | Dose, fractionation and energy consistence between plan and prescription |
| Choice B: | Plan Quality |
| Choice C: | Collision Clearance |
| Choice D: | All above |
| Question 3: 1. Which metric best correlates with time-saving during contouring |
| Reference: | Reference: Femke Vaassen, Colien Hazelaar, Ana Vaniqui, Mark Gooding, Brent van der Heyden, Richard Canters, Wouter van Elmpt, Evaluation of measures for assessing time-saving of automatic organ-at-risk segmentation in radiotherapy, Physics and Imaging in Radiation Oncology,
Volume 13, 2020, Pages 1-6, ISSN 2405-6316, https://doi.org/10.1016/j.phro.2019.12.001. |
| Choice A: | Dice Similarity |
| Choice B: | 95% Hausdorff |
| Choice C: | Added Path Length |
| Choice D: | Mutual Information |
| Question 4: Atlas-based image segmentation requires |
| Reference: | Reference: P. Aljabar, R.A. Heckemann, A. Hammers, J.V. Hajnal, D. Rueckert, Multi-atlas based segmentation of brain images: Atlas selection and its effect on accuracy, NeuroImage, Volume 46, Issue 3, 2009, Pages 726-738, ISSN 1053-8119, https://doi.org/10.1016/j.neuroimage.2009.02.018. |
| Choice A: | large training data |
| Choice B: | optimal template selection |
| Choice C: | training data augmentation |
| Choice D: | intensive convolution operations |
| Question 5: Avoid the following for reproducible application of inference models |
| Reference: | Reference: Matelsky, J., Kiar, G., Johnson, E. et al. Container-Based Clinical Solutions for Portable and Reproducible Image Analysis. J Digit Imaging 31, 315–320 (2018). https://doi.org/10.1007/s10278-018-0089-4 |
| Choice A: | Upgrade to the latest versions of model dependencies |
| Choice B: | Package dependencies used while training in a conda environment. |
| Choice C: | Package dependencies used while training in a Singularity or Docker container |
| Choice D: | Run functional tests on any system change |
| Question 6: AAPM Task Group 275: Strategies for Effective Physics Plan and Chart Review in Radiation Therapy reports that 3 out of the top 10 failure modes for initial plan/chart review of photon/electron EBRT are related to errors in contouring and margins used during treatment planning |
| Reference: | Ref: Ford, Eric, et al. "Strategies for effective physics plan and chart review in radiation therapy: report of AAPM Task Group 275." Medical physics 47.6 (2020): e236-e272. |
| Choice A: | True |
| Choice B: | False |
| Question 7: Current state of the art in auto-segmentation is achieved through: |
| Reference: | Ref: Cardenas, Carlos E., et al. "Advances in auto-segmentation." Seminars in radiation oncology. Vol. 29. No. 3. WB Saunders, 2019. |
| Choice A: | Intensity Thresholding |
| Choice B: | Atlas-based Segmentation |
| Choice C: | Deep learning-based Segmentation |
| Choice D: | All of the above |
| Question 8: Current estimates suggest that AI-based tools will help streamline the treatment planning process, producing more consistent plans and potentially reducing treatment planning errors, providing physicists with additional time for other tasks such as quality improvement, FMEA, research, teaching, and/or physicists-patient consult. |
| Reference: | Ref: Netherton, Tucker J., et al. "The Emergence of Artificial Intelligence within Radiation Oncology Treatment Planning." Oncology 99.2 (2021): 124-134. |
| Choice A: | True |
| Choice B: | False |
| Question 9: Which one of the following is not part of model/epistemic uncertainty |
| Reference: | Kendall, A. & Gal, Y. What Uncertainties Do We Need in Bayesian Deep Learning for Computer Vision? In Advances in Neural Information Processing Systems 30 (eds. Guyon, I. et al.) 5574–5584 (Curran Associates, Inc., 2017) |
| Choice A: | uncertainty in the chosen model structure |
| Choice B: | uncertainty in the estimated model parameters |
| Choice C: | uncertainty due to the lack of training data. |
| Question 10: System meta-analysis had shown that when there was no report of sample level metric, medical ML can be taken with: |
| Reference: | Christodoulou, E. et al. A systematic review shows no performance benefit of machine learning over logistic regression for clinical prediction models. J. Clin. Epidemiol. 110, 12–22 (2019). |
| Choice A: | full confidence |
| Choice B: | caution |
