| Question 1: For non-uniformly sampled Fourier space data in magnetic resonance imaging, fast reconstructions can be performed by interpolation and then inverse fast Fourier Transform. What is the name for the fixed-kernel based interpolation method that results in significantly more accurate images than cubic interpolation: |
| Reference: | Jackson JI, Meyer CH, Nishimura DG, Macovski A. Selection of a convolution function for Fourier inversion using gridding. IEEE Trans Med Imaging. 1991;10(3):473-8. |
| Choice A: | Cartesian Mapping |
| Choice B: | Gridding |
| Choice C: | Bilinear interpolation |
| Choice D: | SENSE |
| Question 2: Magnetic field inhomogeneity disrupts the direct Fourier transform relationship between the image and the data in magnetic resonance imaging. This magnetic field inhomogeneity usually results from the following in and around the brain: |
| Reference: | Truong TK, Clymer BD, Chakeres DW, Schmalbrock P. Three-dimensional numerical simulations of susceptibility-induced magnetic field inhomogeneities in the human head. Magn Reson Imaging. 2002;20(10):759-70. |
| Choice A: | Water. |
| Choice B: | Bone. |
| Choice C: | Air. |
| Choice D: | Muscle. |
| Question 3: Regularization strategies for model-based reconstruction of computed tomography data do NOT include which of the following? |
| Reference: | S. Tilley II, M. Jacobson, Q. Ciao, M. Brehler, A. Sisniega, W. Zbijewski, J. W. Stayman, “Penalized-likelihood reconstruction with high-fidelity measurement models for high-resolution cone-beam CT imaging,” IEEE Transactions on Medical Imaging, 37(4), 988-999 (April 2018) PMID: 29621002, PMCID: 5889122, doi: 10.1109/TMI.2017.2779406. |
| Choice A: | Prior image penalties. |
| Choice B: | Models of system blur. |
| Choice C: | Total variation methods. |
| Choice D: | Learned dictionaries. |
| Question 4: Why does statistical modeling of measurement noise help to improve image quality in CT reconstruction? |
| Reference: | J. Michael Fitzpatrick; Milan Sonka “Handbook of medical imaging, Vol 2. Medical Imaging Processing and Analysis” Chapter 1: Statistical image reconstruction methods for transmission tomography, J. Fessler, 2000 https://doi.org/10.1117/3.831079.ch1 |
| Choice A: | Measurement data can have widely varying signal-to-noise ratio and a statistical model allows relative weighting of the importance of different measurements. |
| Choice B: | There is variability CT motion and this jitter/vibration can be modeled to reduce noise. |
| Choice C: | Statistical models allow one to estimate noise and subtract it from the reconstruction. |
| Choice D: | Noise modeling restricts the reconstruction to use measurements only in a specific attenuation range for more accurate reconstructions. |
| Question 5: How does TOF-PET improve PET image quality? |
| Reference: | “Improvement in Lesion Detection with Whole–Body Oncologic TOF – PET”. El Fakhri G., Surti S., Trott C.M., Scheuermann J., Karp J.S. J. Nucl. Med. 2011; 52: 347-353. |
| Choice A: | Improve spatial resolution. |
| Choice B: | Increase contrast. |
| Choice C: | Increase count rates. |
| Choice D: | Reduce noise especially in large patients or at low contrast. |
| Choice E: | Increase noise. |
| Question 6: Which of the following factors does not affect PET spatial resolution? |
| Reference: | “Physics in Nuclear Medicine" S. Cherry, J. Sorenson, M. Phelps. Saunders Third Edition. |
| Choice A: | Detector size |
| Choice B: | Positron range |
| Choice C: | Patient size |
| Choice D: | Photon non co-linearity |
| Choice E: | Depth of interaction |
