|تاریخ چاپ: 1399/09/06|
|Impact of reconstruction method on quantitative parameters of 99mTc-TRODAT-1 SPECT|
|Author(s):||Tahereh Ghaedian *، Leila Saleki، Fatemeh Ebadi، Abbas Rakhsha|
Quantitative evaluation is recommended to improve diagnostic ability and serial assessment of dopamine transporter (DAT) density scans. We decided to compare the ordered subsets expectation-maximization (OSEM) with filtered back-projection (FBP), and to investigate the impact of different iteration and cut-off frequencies on SBR values.
We retrospectively examined 27 consecutive patients. SPECT reconstruction was performed using OSEM and FBP with Chang’s attenuation correction (AC). Iterative reconstruction parameters were used with different iterations ranging from 2, 4, 6, 8, and 10 with fixed 10 subsets and different subsets including 5, 10 and 15 with fixed 6 iterations. Reconstruction with FBP were performed with different critical cut-off frequencies of 0.3, 0.4 and 0.5.
Comparing SBR derived by OSEM reconstruction with 10 subsets but different iterations revealed statistically significant intraclass correlation (ICC) in both right and left side. There is also no significant difference between different OSEM reconstruction with different subsets and ICC was excellent in all patients. ICC for FBP reconstruction with different cut-off frequency revealed good ICC in all patients. However, lower degree of SBR showed higher decrease in ICC with insignificant and poor correlation in patients with SBR
Our study showed that change in FBP reconstruction parameters can greatly impact the SBR value of 99mTc-TRODAT-1, especially in patients with more severe disease. However, OSEM reconstruction revealed better reproducibility for SBR using different iterations.
|Keywords:||SPECT reconstruction، 99mTc-TRODAT-1، Ordered subsets expectation-maximization، Filtered back-projection|
|Article Type:||Research/Original Article|
|Published:||Iranian Journal of Nuclear Medicine, Volume:28 Issue: 2, 2020|
|Full text:||PDF is available on the website.|