1. ANSWER: B. For the simultaneous acquisition protocol, the resting thal-lium-201 and stress technetium-99m perfusion images are acquired at the same time. This protocol simplifies acquisition and relies on the use of multiple energy windows to acquire the lower-energy thallium-201 at the same time as the higher-energy technetium-99m. A third window is used to correct for the down scatter of the higher-energy technetium-99m into the lower thallium-201 window. The patient is first injected with thallium-201, usually at the normal dose of 3 to 4 millicuries, and then, stress is performed using a lower dose, 9 to 15 millicuries, of technetium-99m. This allows a single acquisition and saves total protocol time. It is not used widely.

2. ANSWER: B. Patient motion in the vertical, horizontal, or rotational planes causes decreased counts in walls that are contralateral or 180 degrees opposite to each other and usually do not follow coronary artery territory distribution. At the edges of the defects, tails or streaming effects may be noted and are responsible for the “hurricane” appearance, which gives this artifact a name. Motion correction may be used to compensate movement, but it can only correct in the vertical plane. Horizontal and rotational motions are much more difficult to detect on the raw projection images and cannot be corrected.

The ramp filter artifact is seen during filtered back projection using a ramp filter and is caused by overlapping or adjacent liver or bowel activity and causes a loss of counts in the adjacent myocardial walls. The use of iterative reconstruction minimizes this artifact but does not make it go away completely. Waiting longer to get liver clearance or administration of small amounts of liquids to enhance peristalsis and move gastrointestinal activity may allow repeat imaging to improve image quality.

Ring artifacts are seen with nonuniformity across the camera head, which may be due to problems with any of the following: photomultiplier tubes, damaged collimator, camera electronics, or sodium iodide crystal. Daily floods are critical to detecting problems at any of these levels.

Skewed artifacts occur when there are errors in the center of rotation, which cause misregistration during reconstruction and a blurring or skewing of the perfusion data.

3. ANSWER: B. The bull’s-eye display (Fig. 4.6) shows severe circumferential reduction in the basal slices for the stress images. There is no combination of coronary lesions that would give this distribution of defects, and this artifact is seen when the base of the left ventricle is incorrectly defined in the lung fields. All the quantitative programs have to deal with hearts of different sizes, and for normal file comparison, hearts have to be expanded or shrunk to a common size. All of the perfusion tracers are taken up by the thick LV walls and there is minimal uptake in normal right ventricles and no uptake in the atria. For that reason, the definition of the basal slices of the left ventricle is critical for accurate quantitative analysis. A first approximation is performed by the computer, but this needs to be visually verified and adjusted.

When the basal plane is put in the left atrium or lung field, quantitative programs shrink the heart to a common size, and when radiotracer is not detected in the basal slices, it assumes that there is severe ischemia when a comparison to normals is performed. When the basal slices are correctly identified on the stress images (Fig. 4.1), quantitative comparison shows that the study is normal. Careful attention is required to define the basal and apical slices, the long axis of the left ventricle, and coregistration between the stress and rest images in order to obtain accurate quantitative analysis. The insertion of the right ventricle is used to register the “clockwise” rotation of the left ventricle. These rules for quantitation are applicable to both PET and SPECT perfusion imaging and are relatively similar for the major software vendors.

4. ANSWER: D. The greater the number of counts in the myocardium, the higher the quality of the images and the more accurate the interpretation. A minimum of 200 counts/pixel using technetium-99m and 100 counts/pixel using thallium-201 in the myocardium on an anterior projection is recommended. Fewer counts will introduce artifacts due to poor performance of the reconstruction filters, which are usually fixed regardless of the total counts acquired. The total number of counts is influenced by the following: body habitus or body mass index, level of exercise or pharmacologic stress, administered dose of radiopharmaceutical, acquisition time, energy window, and the type of collimator. Dose infiltration at the intravenous insertion site and residual activity in the injection syringe may lower the total dose received by the patient. This means that the delivered dose into the patient is less than the measured dose from the dose calibrator.

ASNC Imaging Guidelines and Standards for Single Photon Emission Computed Tomography-2010 http://www.asnc.org/section_73.cfm

5. ANSWER: A. Continuous acquisition allows the gamma camera to acquire counts while it is rotating from one position to the next and increases total counts by eliminating the dead time during camera motion associated with conventional step-and-shoot acquisition. Although there may be image blurring due to the motion, the higher counts and application of appropriate filtering can eliminate the deleterious effects. High-resolution collimation decreases the total counts relative to the use of a general all-purpose collimator. Narrowing the energy window will lower the total counts. ECG gating has minimal effect on counts if separate channels are used to acquire the perfusion data, without applying arrhythmia rejection, and a second channel that acquires the gated information for functional analysis. If only a single channel is used to simultaneously acquire the perfusion and gated information, and arrhythmia rejection is applied, rejected beats will lower the total number of counts in the perfusion images due to beat rejection. This will cause deterioration in the perfusion images, and it is recommended that under these conditions, a very large acceptance window be used to avoid losing counts or prolonging the acquisition if a minimum number of beats or total counts in the myocardium are required.

ASNC Imaging Guidelines and Standards for Single Photon Emission Computed Tomography-2010 http://www.asnc.org/section_73.cfm

6. ANSWER: C. Typically transient ischemic dilatation is caused by severe, multivessel proximal epicardial coronary artery disease (CAD) causing subendocardial ischemia that makes the walls of the left ventricle look thinner and the cavity bigger relative to the resting study. Microvascular coronary disease due to hypertension, diabetes, and renal failure may also result in TID. There may be relatively uniform perfusion when scaled or normalized images are visually interpreted despite a marked decrease in absolute blood flow. Measurements of absolute blood flow may be necessary to identify the presence of diffuse disease.

Severe aortic stenosis may cause subendocardial ischemia in the absence of epicardial or microvascular CAD due to an increase in wall tension with exercise or vasodilator stress, which increases resistance in the coronary circulation. This is accentuated by drops in systemic systolic blood pressure, which decreases the driving gradient across the coronary circulation. Mild aortic stenosis should not cause such adverse hemodynamic changes in coronary blood flow. Mitral regurgitation should decrease LV volume due to an increase in regurgitation fraction and an increase in heart rate, which will shorten the diastolic filling period. Nonischemic cardiomyopathies in the absence of CAD should have increased LV volumes at rest and with stress without visually noticeable changes.

7. ANSWER: C. The number of recommended projections is related to the optimal resolution of the imaging system, which includes the radiotracer. Since technetium-99m images provide higher resolution, the minimum number of projections recommended is 60 to 64. For thallium-201 that has lower resolution, 32 projections are acceptable.

8. ANSWER: B. With 16 frames, there is adequate temporal resolution to accurately assess the 4 phases of diastolic filling and measure peak filling rate and the time to peak filling rate. With 8 frames, there is insufficient temporal sampling of diastole to get reliable information.

With 16 frames, the counts in each frame are half of those obtained with 8 frames. The number of frames does not impact on the spatial resolution of the perfusion data if it is acquired with independent channels for gated and nongated data. When using 8 frames, especially at slower heart rates, the poor temporal resolution often does not capture the true end systolic volume and the EF is underestimated relative to when measuring at a higher temporal resolution. It has been shown with equilibrium radionuclide angiography that 50 to 60 ms temporal resolution is required to get accurate EF measurements.

9. ANSWER: B. Overfiltering SPECT perfusion images, which generally have a low signal-to-noise ratio, makes the images too smooth and leaves too little contrast so that there is a loss of information and a tendency to decrease sensitivity for detection of CAD. Typically, Hanning or Butterworth filters are used with filtered back projection and cutoff frequencies are preset and kept constant for all studies to provide a consistency in the images generated. Ideally, the frequency should vary depending on the actual derived signal-to-noise ratio, but this is seldom done in clinical practice. Ramp filter artifacts are caused by the second filter applied during filtered back projection and are less affected by overfiltering. Filters with cutoff frequencies that vary depending on the actual measured signal-to-noise ratio are available but not in wide use.

10. ANSWER: B. Prone imaging was initially used to get greater separation between the inferior wall of the heart and the diaphragm in order to differentiate an inferior wall infarction from the normal loss in counts caused by the diaphragm, which attenuates to a greater extent than the lung tissue that overlaps the majority of the heart. Prone imaging shifts the location of attenuation relative to supine positioning, and the change in attenuation associated with the change in position can be compared. If a defect is caused by attenuation, the location will shift with a shift in body position while a true infarct will be present regardless of patient position. Subsequent studies have demonstrated that prone imaging may also improve identification of both breast and chest wall fat attenuation by observing the shift in attenuation position. Prone imaging does not correct for attenuation as occurs with CT and line source attenuation correction. For obese patients, the prone position may be uncomfortable and result in greater motion, especially when it is performed after supine imaging. Many facilities shorten the prone imaging acquisition to minimize patient discomfort and subsequent motion, but this does not decrease the total imaging time if it is performed in addition to the supine acquisition. Most sites employing prone images do it selectively in problem patients.