Answer to Question #14000 Submitted to "Ask the Experts"

Can we suppress Compton counts, backscatter, and escape peaks and reduce background in scintillation gamma detectors?

As you may be aware, Compton suppression techniques have been used for some counting applications. The simplest, but not most effective or practical way to increase the peak-to-Compton count ratio is to use a larger volume detector, allowing for relatively more full energy deposition events. Of course, this will likely lead to higher background response from natural and other gamma radiation sources in the vicinity.

Other suppression techniques have been very effectively used when performing gamma spectrometry with solid state detectors, especially germanium detectors. The usual method is to surround the germanium detector with an annular large-volume detector such as sodium iodide or bismuth germanate scintillation detectors or, in some instances, a plastic scintillator. Photons that interact in the primary detector by Compton interactions, such that the scattered photons that escape the germanium detector interact in the annular detector, are rejected through an anticoincidence counting mode. Such Compton suppression techniques will likely also reduce escape peaks since escaping photonic radiation from pair production events within the detector or fluorescent x rays produced near the surface of the detector may deposit some energy in each detector leading to rejection of such events by the suppression system.

Such a technique is generally unsuitable for a gamma camera because of the camera's very large dimensions and special fabrication characteristics. We should note that most modern gamma cameras used in diagnostic nuclear medicine use a process in which pulses from all the individual photomultiplier tubes are summed together to get a signal treated as the full energy deposited in the detector, and if the signal is smaller than what is represented by the full energy deposition of the radiation source gamma ray, the associated events are rejected. This method reduces image degradation associated with photons that might have scattered into the detector from outside the camera.

Compton suppression with an annular detector has major advantages for a typical semiconductor detector vs. a thallium-activated sodium iodide NaI(Tl) detector because the typical sized such gamma detector (usually germanium) has a photopeak detection efficiency considerably less than that for the usual NaI(Tl) scintillation detector, with a commensurately lower photopeak-to-Compton count ratio, especially for moderate to high energy gamma rays. Compton suppression could be feasible for a typical sodium iodide well counter, especially a relatively small volume one, which would benefit more from the suppression than would a larger NaI(Tl) detector. In any case, such a detector would have much smaller dimensions than a gamma camera and is operationally much simpler.

Another technique that has been used for Compton suppression and for suppression of escape peaks has been accomplished by a process described as a sum-coincidence method. For a scintillation detector, such as a well counter, this might also be done by using a second detector (preferably annular detector around the first detector). The electronics must be suited to assessing pulses in both detectors; only when coincident pulses are detected in both detectors the pulses are summed together and the result recorded as an acceptable event. This technique is effective at Compton suppression but it also has the unfortunate defect of reducing full energy deposition events that occur in the first detector, reducing photopeak efficiency.

The backscatter peak results from photons emitted from the sample and scattering from materials that surround the detector and back into the detector. To reduce the magnitude of this peak, the most commonly effective method is to increase the internal dimensions of the shield around the detector. Surrounding the detector with lead, for example, in proximity to the detector may yield appreciable backscatter because the geometry favors interception of shield-scattered photons by the detector. If instead, the detector is housed near the center of a lead cave whose dimensions are much larger than the detector dimensions, the probability of photons from the sample intercepting the shield and scattering back into the detector is reduced.

I hope this is helpful to you.

George Chabot, PhD, CHP