| Question 1: We can precisely predict the optimal dose distribution for a particular patient with a particular beam setup using historical data and deep learning because: |
| Reference: | Nguyen, D., Long, T., Jia, X., Lu, W., Gu, X., Iqbal, Z., & Jiang, S. (2019). A feasibility study for predicting optimal radiation therapy dose distributions of prostate cancer patients from patient anatomy using deep learning. Scientific reports, 9(1), 1076
Nguyen, D., Jia, X., Sher, D., Lin, M.-H., Iqbal, Z., Liu, H., & Jiang, S. (2019). 3D radiotherapy dose prediction on head and neck cancer patients with a hierarchically densely connected U-net deep learning architecture. Physics in Medicine & Biology, 64(6), 065020. doi:10.1088/1361-6560/ab039b |
| Choice A: | Deep learning is a magic that can give you whatever you want; |
| Choice B: | Deep learning can learn the relationship between patient anatomy and optimal dose distribution from the historical data; |
| Choice C: | Optimal dose distributions are all the same for different patients; |
| Choice D: | All of the above |
| Question 2: Do we need to include beam setup information in deep learning models for accurate dose prediction? |
| Reference: | Barragan-Montero, A., Nguyen, D., Weiguo, L., Lin, M., Geets, X., Sterpin, E., & Jiang S. (2018). Three-Dimensional Dose Prediction for Lung IMRT Patients with Deep Neural Networks: Robust Learning from Heterogeneous Beam Configurations. arXiv preprint arXiv:1812.06934. |
| Choice A: | No. All the patients are treated with similar beam setups so no need to worry about beam setups; |
| Choice B: | No. Although sometimes we do have large variation in beam setups, they don’t have a major impact on predicted dose distributions; |
| Choice C: | Yes. When there is a large variation in beam setups we need to include them in the model for accurate prediction in low dose regions; |
| Choice D: | Yes. The more information used the more accurate dose prediction, no matter how relevant the information is; |
| Question 3: Which of the following is true? |
| Reference: | Shen, C., Gonzalez, Y., Klages, P., Qin, N., Jung, H., Chen, L., Nguyen, D., Jiang, S., Jia, X. (2018). Intelligent Inverse Treatment Planning via Deep Reinforcement Learning, a Proof-of-Principle Study in High Dose-rate Brachytherapy for Cervical Cancer. arXiv preprint arXiv:1811.10102 |
| Choice A: | In inverse treatment planning in radiation therapy, the objective function and constraints consist of multiple terms designed for different clinical and practical considerations; |
| Choice B: | The weighting factors of terms in the objective function and constraints are needed to adjusted to achieve an optimal clinical plan with desired trade-offs among various planning objectives; |
| Choice C: | The manual weight tuning process is labor intensive, time consuming, and may lead to sub-optimal final plan quality; |
| Choice D: | Automatic weight tuning is feasible for efficient and high quality treatment planning using deep reinforcement learning; |
| Choice E: | All of above. |
| Question 4: The following represent meaningful applications of AI in medicine: |
| Reference: | Xing L, Krupinski EA, Cai J. Artificial intelligence will soon change the landscape of medical physics research and practice. Med Phys. 2018 Feb 24. doi: 10.1002/mp.12831.
Ibragimov B, Xing L, Deep learning for segmentation of organs-at-risks in head and neck CT images, Medical Physics 44, 547-57, 2017
Thrall JH, Li X, Li Q, Cruz C, Do S, Dreyer K, Brink J. Artificial Intelligence and Machine Learning in Radiology: Opportunities, Challenges, Pitfalls, and Criteria for Success. J Am Coll Radiol. 2018 Mar;15(3 Pt B):504-508. doi: 10.1016/j.jacr.2017.12.026. Epub 2018 Feb |
| Choice A: | Image segmentation and classification |
| Choice B: | Image analysis and disease diagnosis |
| Choice C: | Clinical decision-making |
| Choice D: | Treatment planning |
| Choice E: | All of the above. |
| Question 5: Deep learning algorithm has the following features: |
| Reference: | Wu Y, Ma Y, Liu J, Du J, Xing L, Self-attention convolutional neural network for improved MR image reconstruction, Information Sciences 490, 317-328, 2019.
Ibragimov B, Toesca D, Yuan Y, Koong A, Daniel C, Xing L. Neural networks for deep radiotherapy dose analysis and prediction of liver SBRT outcomes. IEEE J Biomed Health Inform. 2019 Mar 11 Preview Abstract PMID: 30869633.
Yuan Y, Qin W, Buyounoski M, Hancock, S, Han B, Xing L, Prostate Cancer Classification with Multi-parametric MRI Transfer Learning Model, Med Phys, 46, 756-765, 2019. PMID: 30597561. |
| Choice A: | It uses a large amount of annotated datasets to train a deep learning model |
| Choice B: | It is generally computationally intensive and done using GPU |
| Choice C: | Overfitting may happen if the training data is not sufficient |
| Choice D: | It extracts automatically extracts useful features to build a model |
| Choice E: | All of the above. |
| Question 6: Response adaptation is based on |
| Reference: | The Role of Machine Learning in Knowledge-Based Response-Adapted Radiotherapy. Tseng HH, Luo Y, Ten Haken RK, El Naqa I. Front Oncol. 2018 |
| Choice A: | CT imaging only |
| Choice B: | Blood-based cytokines |
| Choice C: | CT imaging and blood-based biomarkers |
| Choice D: | All available knowledge (panomics) |
| Question 7: Decision support systems for radiotherapy requires |
| Reference: | Reference: Prospects and challenges for clinical decision support in the era of big data. El Naqa I, Kosorok MR, Jin J, Mierzwa M, Ten Haken RK. JCO Clin Cancer Inform. 2018. |
| Choice A: | IBM Watson oncology |
| Choice B: | Natural language processing tools |
| Choice C: | Outcome models and optimization scheme for decision making that can be developed using machine learning |
| Choice D: | Better statistical packages |
| Question 8: Reinforcement learning |
| Reference: | Reinforcement Learning: An Introduction. Sutton, RS.; Barto, AG. MIT Press,1998. |
| Choice A: | Machine learning algorithms that can take a proper action given a defined environment and a suitable reward function |
| Choice B: | A psychology principle about learning. |
| Choice C: | A deep learning algorithm. |
| Choice D: | None of the above. |
| Question 9: What� the optimal images for radiation therapy treatment planning? Using CT simulation images as an example. |
| Reference: | Dolly S., Anastasio M., Li H., "Task-Based Image Quality Assessment in Radiation Therapy: Initial Characterization and Demonstration with CT Simulation Images", SPIE Medical Imaging Conference Proceedings, 2017, Proceeding Volume 10136. Oral Presentat |
| Choice A: | Images acquired with minimal dose. |
| Choice B: | Images acquired with as maximum as possible radiation dose allowed by an CT scanner. |
| Choice C: | Images acquired with the scanning parameters optimized based on the performance of target contouring methods. |
| Choice D: | Images acquired with the scanning parameters optimized based on the treatment outcome. |
| Choice E: | Does not matter, CBCT image quality is more important |
| Question 10: Which applications that therapeutic operating characteristic (TOC) curves can be employed for? |
| Reference: | Barrett, H. H., Myers, K. J., Hoeschen, C., Kupinski, M. A., and Little, M. P., "Task-based measures of image quality and their relation to radiation dose and patient risk," Physics in Medicine & Biology 60(2), R1 (2015). |
| Choice A: | Comparison of different imaging instruments |
| Choice B: | Comparison of OAR and target delineation, and image registration algorithms |
| Choice C: | Comparison of RT treatment planning algorithms |
| Choice D: | Determination of optimal treatment dose for individual patients when biological accuracy can be clearly established. |
| Choice E: | All of the above. |
