Decaying source method for scintillation camera resolving times.

An earlier paper by the author showed that scintillation camera systems can be described effectively using the resolving times T and tau0 of the dominant nonparalyzable and paralyzable components, that is, the detector system and the computer interface, respectively. When used with a full spectrum window, the camera has a lower nonparalyzable and an upper paralyzable operating range with normalized threshold input rate n(t)= N(t)tau(0) = ln(1 +k(T)n(t)), where kT= T/(tau)0. Correct determination of T and tau(0) requires that both r12 and r in the normalized two-source equations kT=(2/r12- 1/r) and k0=(r12/2r2)ln(2r/r12) come from the nonparalyzable (n< or =n(t)) and paralyzable (n> or =n(t)) ranges, respectively. A serious constraint of the two-source method, therefore, is that the large ratio a = n12 ln = 2 can lead to an input rate range (n,n12), which includes the threshold point n,, and in which neither T nor tau0 can be measured correctly. The decaying source method constitutes a refinement of the two-source method, which enables smaller ratios 1<a<2 to be used, and also includes the two-source method as a special case (a=2). This new method requires just two consecutive readings on a single decaying source, as opposed to three measurements on two sources of activity, for each determination of T or tau0, thus also minimizing staff exposure. The fact that only count rates and time intervals need now be recorded greatly simplifies computerization of the data acquisition and analysis activities, and the potential for real-time applications is obvious. The method enables T and tau0 to be measured accurately and with sufficient resolution to reveal possible variations with input rate. Long measurement times using a decaying source can be avoided, if required, by using a set of decaying sources simultaneously to cover different portions of the count rate range. The application of the measurement procedure in real-time and the use of the resolving times in the accurate correction of deadtime losses, also in real-time, are treated separately in another paper.