LAUSR

Cardiac imaging with positron-emission tomography (PET) affords non-invasive measurements of myocardial blood flow (MBF) and myocardial perfusion reserve (MPR), improving traditional nuclear medicine in diagnosis and risk stratification of patients with coronary artery disease (CAD). These measurements can be obtained in few minutes with appropriate software packages by applying tracer kinetic modeling to dynamic PET images. Any numerical value that we receive from dynamic cardiac PET results from this transformation. It should be considered that several technical aspects can affect MBF and MPR measurements, including physical and biochemical properties of different tracers and specific features of PET instrumentations. For clinic applications, Rubidium (Rb) is the most widely used tracer, considering that it is generator-produced and it does not require a cyclotron in site. Several software tools for absolute quantification of dynamic PET with Rb have been developed and a comparison between those commercially available software programs have been already performed. In particular, Nesterov et al compared 8 tools on the same set of data and they found that different software provided similar results when the same kinetic model was applied. However, it should be also taken into account that quantitative analysis of cardiac PET imaging is a challenging process and kinetic model is the latest in a series of steps that can affect the results, including reconstruction, segmentation, quality control, and outputting the values. In particular, segmentation is a crucial step that precedes kinetic modeling and allows to localize the left ventricular (LV) myocardium and the blood in its cavity in order to obtain time activity curves (TAC). For this purpose, regions of interest (ROI) of the LV cavity and myocardium can be used. Definition and placement of ROI, as well as reorientation process from transverse images into a standard cardiac orientation and differences in tissue sampling between rest and stress series, may need more than one of the manual interventions and affect the results when different software packages are considered. Semiautomatic and automatic segmentation methods, based on various statistical scenarios, have been proposed to improve the reproducibility of PET quantification. Factor analysis is widely used for extracting tissue TAC in dynamic PET images. This approach is based on the assumption that dynamic PET noise and the model approximation errors follow Gaussian distributions. Differently, non-negative matrix factorization (NMF) uses Poisson statistics as noise model. But these statistical approaches were developed primarily with an assumption of homogeneous radiotracer uptake in the same organ and they do not fully take into account the partial volume effect commonly exists in dynamic cardiac PET. More recently, a triple-factor non-negative matrix factorization (TNMF) method for semiautomatic LV cavity and myocardial segmentation has been proposed and it has been found to be highly feasible for MBF and MPR quantification. In the current issue of Journal of Nuclear Cardiology, Liu et al proposed a new software tool called ‘‘Yale-MQ,’’ based on an optimized TNMF segmentation workflow and on one-tissue compartment model. The authors evaluated quantitative precision and intra-/ inter-observer variabilities of Yale-MQ in MBF and Reprint requests: Roberta Assante, MD, PhD, Department of Advanced Biomedical Sciences, University Federico II, Via Pansini 5, 80131 Naples, Italy; [email protected] J Nucl Cardiol 2021;28:1249–51. 1071-3581/$34.00 Copyright 2020 American Society of Nuclear Cardiology.