Generation of Blood Volume and Myocardium Images

The quantification of cardiac PET studies requires that the tracer uptake and clearance in the myocardium is monitored by a dynamic acquisition sequence starting at the time of tracer application. To adequately sample the rapidly changing activity concentrations the acquisitions need to be very short at the beginning, and may last somewhat longer at later times after the bolus has passed the heart for the first time. As a consequence of the short acquisitions, the individual images are noisy and often do not show enough anatomical detail to

• reorient the images into the standard short axis orientation, and to
• localize the different heart segments and the blood blood spaces.

PCARD provides two mechanisms for generating images with more anatomic information from the dynamic sequence:

1. Average Subtraction Method: This approach is simple and universal. It is based on the fact that a tracer which is applied as a venous bolus first arrives with a high concentration in the right ventricle, after the passage through the lungs with somewhat decreased activity in the left ventricle, and only afterwards reaches the myocardium, while the concentration in the blood spaces decreases by the continuous clearance.
   Therefore, an image of the blood volume space can be generated by averaging some of the very early acquisitions. The optimal begin and end times of the averaging depend on how fast the bolus was applied, when the PET scan was started, and must therefore be adjustable for each acquisition.
   Similarly, tracers tend to accumulate in the myocardium (except for water) over time. Therefore, averaging a range of late frames can potentially provide an anatomical image of the myocardium. To improve the myocardium contrast PCARD supports the option to subtract from the myocardium images a weighted fraction of the blood volume images.
   The optimal parameters used for the average subtraction method must be experimentally determined. Once a good setting has been found it should be usable for studies of the same type as long as the tracer application is done in the same way and acquisition protocol is constant.
2. Factor Analysis (Water only): The factor analysis is a more sophisticated approach to derive weighting factors for the different acquisitions. A blood image is obtained by averaging all frames scaled with the respective blood factors, and a myocardium image by averaging all frames scaled with the respective myocardium factors.
   The factor analysis is model-driven. It requires an approximate input curve, a typical myocardium blood flow value (depending on the condition), and time delays between the occurrence of the activity in the ventricles and the myocardium. Given this information, the expected activity in myocardium can be calculated. Together with the blood activity it is fed into a factor analysis procedure which calculates the factors which provide an optimal contrast between the blood and the myocardium.