THEORETICAL INVESTIGATION OF THE ESTIMATION OF RELATIVE REGIONAL NEURORECEPTOR CONCENTRATION FROM A SINGLE SPECT OR PET IMAGE

The validity of estimating changes in regional neuroreceptor concentration based upon a single ECT image was examined. The focus of this problem is to determine whether an image can be acquired at a time when changes in the observed regional radioactivity are much more sensitive to changes in receptor concentration than to changes in radioligand delivery. These sensitivities were defined as the normalized partial derivatives of the regional radioactivity signal (S) with respect to the total receptor (R,) and with respect to radioligand delivery (k1). Using computer simulations, we find that (dS/dR,)/(S/R,) can be less than unity (receptor hyposensitivity), approximately equal to unity, or significantly greater than unity (receptor hypersensitivity). The degree of receptor sensitivity depended on the values for the model parameters, as well as on the time at which the image was acquired and the specific activity of the radioligand. It was possible to find classes of parameter sets under which the receptor sensitivity was close to unity and the delivery sensitivity was of much lower magnitude. These results indicate that, if the parameters for a given radioligand-neuroreceptor system can be established as belonging to one of these classes, then changes in regional neuroreceptor concentration can be estimated based upon a single ECT image. © 1990 IEEE