Despite the disadvantages, if there is a long list of patients to image then dual energy acquisition can be helpful. If there are only a few patients then single alternate views give the best images. If you want to use dual acquisition then you should carry out some checks.

The main issue is downscatter into the Tc99m window. This will depend primarily on the generator strength but other factors are the energy windows used and the breathing system used. The contribution of downscatter to the Tc99m images should be checked for your camera system and generator. However first know your camera!

Camera Questions

1. How does your camera operates in dual energy mode?

2. What are the stop conditions?

3. If a count based stop condition is used, is it the sum of the counts in both channels or the counts in the first channel ?

4. Is it possible to separately control the stop condition of each window?

5. Check that you do get separate images at the end of the acquisition!

Happy that you now know how your camera does dual-acquisition? Good. Now do some single and dual-acquisition tests to calculate the downscatter component.

1. Carry out a single view study of a patient, except collect the Kr81m views on dual acquisition. Only one patient position is required and the posterior view is recomended. As long as the Kr81m counts are the appropriate value required (you may need to check your stop conditions) then this does not add in any way to the time or radiation dose for the patient. This gives both a single Tc99m view without simultaneous Kr81m and also a Tc99m view acquired with Kr81m present.

2. Determine the total counts and time for each view.

3. Scale the single acquisition Tc99m view counts to the time of the dual acquisition study

4. Subtract the scaled counts from the dual acquisition Tc99m counts and express the answer as a percentage of the scaled counts.

5. Note the Kr81m generator activity and calculate the activity which corresponds to the downscatter value you will accept (20% should be OK and some departments accept up to 30%)

example

Assume you calculate the generator has 200MBq Rb81/Kr81m when you image the patient. Also assume you normally collect 400k images for Tc99m and Kr81m.

1. posterior single Tc99m view -           400,000 counts in 90s

2. dual acq. posterior Tc99m view -     500,000 counts in 100s
   (Kr81m acquisition is 400,000 counts in 100s)

3. scale the single Tc99m view -   444,444 counts in 100s

4. subtract from the dual acq counts
   500,000 - 444,444 = 55,556

5. express as a percentage of the single count
   55,556 is 12.5% of 444,444

6. The downscatter in this case = 12.5%

7. If you accept 25% downscatter then you can image with a Kr81m generator with up to twice the activity ( 400MBq in this example)

8. To keep 400k Tc99m counts in the perfusion image, you will need to collect up to 511k counts in the dual Tc99m image.
   400,000 + 2x55,556 = 511,112 counts

9. A complex stop condition can result

In this example 12.5% of the Tc99m image counts are due to downscatter from the Kr81m in dual acquisition mode . In fact , this value is generally acceptable. A value approaching 25% may even give acceptable Tc99m perfusion pictures. This implies that a generator strength of double the activity might be used in dual acquisition mode.

However a generator with twice the activity means that the Kr81m view will finish (ie 400k)  in 50s. This time is too short for the Tc99m view. Therefore the stop condition needs to be set to allow the perfusion image to be collected long enough (90s in the above example). This will mean that the Kr81m image may also have to be collected for up to 90s unless separate stop conditions are possible.

With the above conditions it is always possible to stop the air flow to the generator and remove the patient mask after the 50s but continue to collect the dual energy study until the time of 90s for the Tc99m view. This will in fact cut down the downscatter as well. The main problem is ensuring that the patient does not move when the mask is removed.

Some departments acquire the posterior views first separately on single acquisition. This gives a guide to Kr81m and Tc99m times for the other views to be counted in dual acquisition mode.

In the above example, If the generator activity on another day happened to be  much greater than 400MBq then it would be important to stop the Kr81m view early in the dual acquisition study.

Of course, if you have a high activity Kr81m generator then why bother with dual energy acquisition? The Kr81m images will be very quick, so why not just collect single views? The images will be better quality and you avoid the risk of too much downscatter.

The final issue is that later in the day or with a low activity generator then it may become important to use dual energy acquisition to speed things up. But make sure the Kr81m view now has enough counts with any stop condition that is set.

With all the above issues, a patient protocol will depend on local circumstances and equipment. However we have given a typical protocol for information