CPET - Simultaneous Transmission/Emission considerations

Scenario - Whole body torso scan (6 beds x 10.8[cm/bed] FOV) of adult male where:

  1. transmission scan is performed on adult subject (T) 5[min/bed]
  2. subject is injected with 10[mCi] 18-FDG and a second transmission scan (T+E) of 5 [min/bed] is conducted during uptake phase
  3. regular whole body emission image is collected at 7 [min/bed]
  4. attenuation files are created using the (T+E) and T scans and all bed positions are reconstructed using both T and (T+E) attenuation files.

ECAT specific considerations

In protocols where the transmission sources are extended, the lower and upper energy windows used to discriminate true 511 [keV] photons from scatter and pileup are set to 250 and 650 [keV] respectively. In protocols where emission data is collected, the upper energy window is extended to 850 [keV]. In a (T+E) scenario where the narrower window is used, more emission events will be rejected than is observed in a typical emission study (where the wider window is used and transmission events aren't present). Data to be displayed on how significant this loss is once we've collected it.

(T+E)/T ratio of total plane net trues

[plot 8,104 bytes] - This plot shows the ratio of net trues in each plane for the (T+E) transmission scan divided by the T transmission scan. Note that this number is slightly > 1.0 for most planes indicating the emission data does contribute to the transmission data; however, this contribution (in terms of whole plane counts) is minimal, being on the order of a 1+/- 3% increase in overall counts.

(T+E)/T ratio for LOR attenuation coefficients

[plot 4,935 bytes] - This plot shows the ratio of attenuation correction factors for the (T+E) scan divided by those computed from the T scan alone for all Lines Of Response (LORs) across all bed positions. When emission data is present during a standard transmission scan, the system is lead to believe that attenuation along a LOR crossing emission sources is less than it actually is. When this occurs, the ratio of (T+E)/T will be less than 1.0. Note that is evident (from the area under the curve) that there are a significant number of pixels for whom an incorrect attenuation correction is computed in the (T+E) scenario, and many of these LORs will experience more than a 10% error (where (T+E)/T < 0.9) in their attenuation correction factors.