2022 AAPM 64th Annual Meeting
Back to session list

Session Title: Integrating Omics in the Era of AI for Better Patient Specific Outcomes
Question 1: Multi-omics refers to:
Reference:Machine and Deep Learning in Oncology, Medical Physics and Radiology, I El Naqa and M Murphy (Eds), Second edition, Springer-Nature, 2022.
Choice A:Biological markers (genomics, proteomics, transcriptomics, etc)
Choice B:All large-scale data from imaging (radiomics), pathology (pathomics), dosimetry (dosiomics) , biology (genomics, proteomics, transcriptomics, etc)
Choice C:Another type of biotechnology data extracted from gene expression
Choice D:Machine learning algorithm
Question 2: Radiomics represents a method for the quantitative description of medical images.
Reference:van Timmeren, Janita E., et al. "Radiomics in medical imaging—“how-to” guide and critical reflection." Insights into imaging 11.1 (2020): 1-16.
Choice A:True
Choice B:False
Question 3: Maximum likelihood estimation gives more stable results compared to least square error methods when fitting normal tissue complication probability models.
Reference:Moiseenko, Vitali, Lawrence B. Marks, Jimm Grimm, Andrew Jackson, Michael T. Milano, Jona A. Hattangadi-Gluth, Minh-Phuong Huynh-Le, Niclas Pettersson, Ellen Yorke, and Issam El Naqa. "A primer on dose-response data modeling in radiation therapy." International Journal of Radiation Oncology* Biology* Physics 110, no. 1 (2021): 11-20.
Choice A:True
Choice B:False
Question 4: Which of the following is not an appropriate validation scheme?
Reference:Transparent reporting of a multivariable prediction model for individual prognosis or diagnosis (TRIPOD): the TRIPOD Statement. Collins, G.S., Reitsma, J.B., Altman, D.G. et al. BMC Med 13, 1 (2015)
Choice A:Random split the dataset into training/test groups, develop the model on the training dataset and test the model on the test dataset
Choice B:Develop a model on an internal institutional dataset and test the model on an external institutional data
Choice C:When many features are available, pre-select the variables using the whole dataset, and then develop and test the model using cross-validation.
Choice D:Split the dataset into training/test groups based on time, develop the model on the training dataset and test the model on the test dataset
Question 5: Which of the following is not a method to improve the interpretability of the model?
Reference:Balancing accuracy and interpretability of machine learning approaches for radiation treatment outcomes modeling. Luo, Y., Tseng, H. H., Cui, S., Wei, L., Ten Haken, R. K., & El Naqa, I. (2019). BJR open, 1(1), 20190021.
Choice A:Grad-CAM
Choice B:LIME
Choice C:CART
Choice D:Deep Learning
Question 6: When starting a radiomics study one should:
Reference:Ger, Rachel B., et al. "Comprehensive investigation on controlling for CT imaging variabilities in radiomics studies." Scientific reports 8.1 (2018): 1-14.
Choice A:Resample the image to a consistent voxel size and include all reconstruction algorithms
Choice B:Resample the image to a consistent voxel size and include only similar reconstruction algorithms
Choice C:Resample the image to a consistent voxel size, include only similar reconstruction algorithms, and only include patients with the same mAs setting
Choice D:Include all patients as is
Back to session list