Answer to Question #1381 Submitted to "Ask the Experts"

Q

I know that the half-life of ^99mTc is six hours. I start with 40m Ci of activity; 12 hours later, I would have 10m Ci. If I am using a low-energy, high-resolution (LEHR) collimator and the absolute system sensitivity is > 250 cpm/uCi, how many counts would be detected by the gamma camera in five minutes? I also know that the maximum count rate is > 240 kcps, the specs are given for one head, and I am using two heads. I calculated 25,000,000 counts. My question is how this, the number of counts detected, is related to the total number of counts in the vial.

A

When determining the absolute system sensitivity (or count per emitted photon efficiency) with a nuclear medicine gamma camera, or other sodium iodide (NaI) crystal, there are a number of parameters to consider. These would include such things as source-detector distance, size and geometry of both, attenuation of photons (that is, with a vial, or patient and collimator), resolving time, intrinsic efficiency of the NaI crystal at a given photon energy, etc. To answer the question, you must first determine what you want to know, for example, efficiency of ^99mTc photon detection at three feet from a small 25 ml vial, with a given collimator. (Note: I'm not sure what this LEHR collimator is, but if the collimator has some focusing property where you're going to compare this vial calibration to a patient organ content measurement, you may want to choose a similar setup.) Regardless, having a source-detector distance of four to five times the maximum length of the vial will provide a condition where the vial looks like a "point source," and the inverse square relationship will hold. Also, the calibration setup should have minimal photon scattering, so holding the vial secure, in air, a few feet off the floor should be an experimental goal.

As you noted, ^99mTc has an approximate 6.02 hour half-life. To get an "absolute" calibration, use a nuclear medicine well ionization chamber dose-calibrator (which should be calibrated and accurate) to draw up a test sample. If this is to be a typical patient "dose," you could draw that 40 millicuries (mCi) of ^99mTc. Hold it in the desired geometry and repeatedly do a five-minute count over a few days. A plot of this data (in say counts per second) on semilog paper or with a spreadsheet will illustrate any resolving time concerns. If the data is linear through the entire test period showing an exponential decay, there is no significant count loss. If nonlinearity is noted in the initial counting phase, then note the time where it becomes linear, use one of the later data points, and from the well chamber "exact" initial mCi—calculate the vial activity at that time. Also, convert from mCi to disintegrations per second. This information can now be used to calculate the actual system efficiency in counts per disintegration (c/d). This number should be less than 0.50 c/d for a "two pi" geometry experimental setup and, unless you have the source right up to the camera face, should be quite a bit lower than 0.50 c/d.


David J. Allard, CHP