daily clinical applications, utility is very frequently defined by physiological and

methodological variability rather than systematic error (bias).

16.4.3.1 Limitations of the Retention Model

Principal limitations of the retention model are:

•

One assumption made in this model is that there is no wash-out of tracer. Though

this approximation holds good for only normally perfused, feasible myocytes,

nonetheless tracer eventually gets washed out in the case of severe ischemia or

non-transmural scar.

•

Assumption that the entire integrated arterial input function can be captured

during the initial fixed 2-min blood pool image may not hold good in case of

some physiological delay or low heart function.

•

Evaluation of the partial volume correction factors used in the model should be

done with the help of phantom scans for each PET scanner and radionuclide.

16.4.4 Comparison Between Retention Model and Compartment

Model

A comparison of these two models can be made as follows:

•

The short-term repeatability expressed as 95% repeatability coefficient, RPC, is

found to be 15–20% for retention model and about 20% for compartmental

modeling.

•

Plot in Fig. 16.8 shows Bland-Altman global left ventricular flow data depicting

repeatability limits of 95% (RPC  20%, dashed lines) (R-flow simplified

retention model, C-flow compartment model). 27% of rest scans and 32% of

stress scans were found to lie outside the repeatability limits (Moody et al. 2020).

•

Thus, it signifies a prominent lack of agreement in about 33% of patients

considered.

16.5

Software Model and Tools

Based on simplified retention model, the following commercial and academic

software packages are available:

1. HeartSee: This software has been developed by the University of Texas,

Houston. This software used for cardiac positron emission tomography (PET)

determines both local and global complete rest and stress myocardial perfusion. It

maps in patients with suspected or known coronary artery disease (CAD) and

hence is useful in quantifying clinical interpretation of PET perfusion images and

hence predicting their severity.

16

Role of Microfluidics and Nanofluidics in Managing CAD

285