Lost in quantification…: The influence of different software packages on flow quantification measures

Abstract

Over the last decades, non-invasive myocardial perfusion imaging has shown major improvements and has consolidated its role as a mainstay for the workup of coronary artery disease. 1 While the most often applied technique for myocardial perfusion imaging is still single-photon emission tomography (SPECT) imaging, positron emission tomography (PET) for myocardial perfusion imaging has gained growing importance and acceptance in clinical routine. Reasons for that development are-besides the persisting intermittent shortage of Technetium-99m-technical advantages such as reliable tracer uptake quantification, higher count sensitivity, higher temporal and spatial resolution of PET. 2 Furthermore, PET myocardial perfusion imaging gets along with a substantially lower radiation exposure to the patient due to different physical properties of PET radiotracers such as shorter half-life. 3 Last not least, the probably most important advantage of PET myocardial perfusion imaging is the potential for absolute quantification of myocardial blood flow, a technique which has first been described in the 80s. 4 The superior extraction fraction of PET perfusion tracers compared to Technetium-99m-labeled SPECT agents and the calculation of absolute measures of myocardial blood flow also are main reasons for the superior diagnostic accuracy of PET over SPECT. [5][6][7] However, absolute quantification of myocardial blood flow is not trivial and there are many factors that may have major impact on the calculated results. 8 As indicated before, the different available PET perfusion tracers also have different physical (half-life, positron range etc. of the respective radionuclide) and biochemical [extraction/retention fraction, excretion mechanism, accumulation (e.g., N-13 ammonia, Rubidium-82) vs diffusion (O-15 water) etc.] properties, which all have to be taken into account. Also, the protocol needs to be designed carefully: for example, the generation of an accurate input function, which is calculated from a small region of interest (ROI) in the left atrium, left ventricular cavity, or the aorta when the radiotracer passes the heart may be incorrectly assessed if the radiotracer is injected too fast (i.e., with a too high activity concentration) which may drive the scanner into saturation. 9 The ROI location for the estimation of the input function per se may also have a substantial effect on blood flow results. 10 Obviously, motion and subsequent misalignment between PET and CT data may cause erroneous estimation of the myocardial blood flow. 11 It goes without saying that different scanners with different properties (crystal material, 2D vs. 3D scanners, timeof-flight availability etc.) may result in different estimations of myocardial blood flow. 12 Another important factor that may have substantial impact on myocardial blood flow estimation is the choice of kinetic model, which has to be picked carefully and in dependence of different factors, e.g., the applied perfusion tracer. 13 Another important factor (also on the post-pro...