Aim: The aim was to study the influence of single-photon-transmission scan duration in 3D-PET on the quantitative values of attenuation coefficients and noise in transmission images and of activity concentrations and noise in attenuation corrected emission images of thorax phantom- and patient data.
Method and material: Using dual collimated Cs-137 singles transmission sources (E gamma = 662 KeV, A = 2* 614 MBq) on an ECAT ART tomograph series of transmission scans of a thorax phantom were acquired pre- and post-injection of 18F. 17 patients underwent two transmission scans. The scan time of the short transmission was chosen according to the results of the phantom studies (noise of Poisson statistics less than 4%). Transmission and attenuation corrected emission images were evaluated with respect to estimated linear attenuation coefficients, noise and specific activities in organs.
Results: The phantom studies reveal little variation of the estimated linear attenuation coefficients as a function of scan duration. The estimates of the attenuation coefficients are found to be 1% lower than the expected values for pre- and up to 6.5% lower for post-injection transmissions. The noise level in the transmission images increases as expected for Poisson data. The noise level in the attenuation corrected emission images shows only little increase with decreasing transmission scan time whereas it is strongly influenced by a variation of emission scan time. In patient studies, less than 3% difference was found in the estimated linear attenuation coefficients as well as in the activity concentrations between short (pre or post-injection) and long transmission scans. The noise levels in transmission and emission images are 1% (pre-injection) and 4% (post-injection) higher for short transmission scans.
Conclusion: Due to the high photon flux, single photon transmission offers good clinical performance with significantly reduced transmission scan durations (< 2 min/bed in pre-, < 4 min/bed in post-injection transmission).