Nuclear Medicine Physics: Collimators Part 2
Review of nuclear medicine collimators for radiology board exams including the ABR core exam. Brought to you from your favorite radiology podcast for board review.
Remember collimators have a lot to do with the resolution of an imaging study. The way you get positional data on an image through collimators is that a collimator only accepts counts from a small region of the object being imaged that is located directly in front of the collimator hole. The smaller the collimator holes, the higher the resolution, but also the lower the sensitivity because you will capture fewer counts in each amount of time.
For illustration (I don’t think you don’t have to memorize these values):
-LEAP accepts about 500k counts per minute whereas LEHR accepts about 175k counts per minute. So, you have lower sensitivity with a LEHR than a LEAP when imaging for the same amount of time with the same injected activity. But because the collimator holes are smaller and deeper on a LEHR, the LEHR has better resolution than a LEAP because it accepts fewer scattered gamma rays.
-LEAP has resolution of about 10 mm at 10 cm from the collimator, whereas a LEHR has a resolution of about 6 mm at 10 cm from the collimator.
What about Iodine 123? Several studies have shown improvement in image quality when using a medium rather than low energy collimator. I don’t think they will ask this on the ABR Core Exam though because some sites may vary in what collimator is used.
Basically, on a multiple-choice exam, if most of the energy is 140 keV or lower I would choose to use a low energy collimator (remember I123 has 159 keV), if you are using I131 choose high energy, and otherwise go with medium energy.
What collimator do you use for the weekly bar phantom spatial resolution QC test? A low energy high-resolution collimator.
Collimators are made of lead or tungsten—you need dense metals so the septa can be made as thin as possible.
For nucs QC testing an “extrinsic” test has the collimator in place. An “intrinsic” test is done without a collimator. I remember this by thinking that an “extrinsic” test has the radioactive source “external” to a collimator. An “intrinsic” source tests the intrinsic performance of a gamma camera without an external collimator.
How do we get rid of scatter in nuclear medicine? A few methods to be aware of:
1. Collimators
a. Counts that come in off-axis will get blocked by the septae of the collimator and not reach the crystal face
i. Remember the increasing the effectiveness by which the collimator blocks scatter will result in a lower sensitivity
ii. 2D PET had collimators (tungsten, not lead), 3D PET does not have collimators. Hence 3D PET is more sensitive and uses other methods such as time-of-flight to reduce scatter.
b. Pulse-height analyzers
i. Selecting a narrow energy range on the pulse-height analyzer will make it so the system only uses gamma rays with the desired energy for an image
ii. Scattered gamma rays that have given off portions of their energy from the Compton scatter event/collision will not be accepted in the final image
iii. If you select the wrong energy, you get low-resolution images with poor anatomic localization
1. This is called “off peak” imaging
2. This may happen if a technologist forgets to switch from a peak of 140 keV for Tc99 when imaging with something that has a different energy like Ga67.
3. Also, commonly may forget on first patient of day to switch from Co57 peak (122 keV) (used for cobalt flood QC testing) when imaging with Tc99m (140 keV) or another radiopharmaceutical.
c. Time-of-flight and coincidence detection on PET scanners
i. Will cover in more detail on discussion of PET physics
ii. Uses a PET detector ring that has such high temporal resolution that it can detect within nanoseconds if a photon pair strikes one detector slightly before the other 180 degree opposed detector
iii. Can use this difference in photon pairs hitting detectors to figure out the event must have occurred closer to the detector that detected the event a few nanoseconds earlier
iv. By using this method, you can reduce the length of the line of response and thereby get rid of scatter and improve resolution of an image
v. Coincidence detection also reduces scatter on its own because you only accept counts that are spatially correlated. The system will only accept counts that are registered at the same time on detectors that are 180 degrees opposed—therefore assumed to represent the non-scattered annihilation photon pair.
Collimator
keV
Radiopharmaceuticals
LEHR/LEAP
140 and below
Tc99m, Thal201
Medium Energy
Between Tc and I131 (140-365 keV)
Ga67, In111, I123
High Energy
364 keV and above
I131
