Evaluation of iterative reconstruction method and attenuation correction on brain dopamine transporter SPECT using anthropomorphic striatal phantom

Maebatake, Akira and Imamura, Ayaka and Kodera, Yui and Yamashita, Yasuo and Himuro, Kazuhiko and Baba, Shingo and Miwa, Kenta and Sasaki, Masayuki (2016) Evaluation of iterative reconstruction method and attenuation correction on brain dopamine transporter SPECT using anthropomorphic striatal phantom. Asia Oceania Journal of Nuclear Medicine and Biology, 4 (2). pp. 72-80.

AOJNMB_Volume 4_Issue 2_Pages 72-80.pdf

Download (1MB) | Preview
Official URL: http://aojnmb.mums.ac.ir/article_7007.html


Objective(s): The aim of this study was to determine the optimal reconstruction parameters for iterative reconstruction in different devices and collimators for dopamine transporter (DaT) single-photon emission computed tomography (SPECT). The results were compared between filtered back projection (FBP) and different attenuation correction (AC) methods.Methods: An anthropomorphic striatal phantom was filled with 123I solutions at different striatum-to-background radioactivity ratios. Data were acquired using two SPECT/CT devices, equipped with a low-to-medium-energy general-purpose collimator (cameras A-1 and B-1) and a low-energy high-resolution (LEHR) collimator (cameras A-2 and B-2).The SPECT images were once reconstructed by FBP using Chang�s AC and once by ordered subset expectation maximization (OSEM) using both CTAC and Chang�s AC; moreover, scatter correction was performed. OSEM on cameras A-1 and A-2 included resolution recovery (RR). The images were analyzed, using the specific binding ratio (SBR). Regions of interest for the background were placed on both frontal and occipital regions.Results: The optimal number of iterations and subsets was 10i10s on camera A-1, 10i5s on camera A-2, and 7i6s on cameras B-1 and B-2. The optimal full width at half maximum of the Gaussian filter was 2.5 times the pixel size. In the comparison between FBP and OSEM, the quality was superior on OSEM-reconstructed images, although edge artifacts were observed in cameras A-1 and A-2. The SBR recovery of OSEM was higher than that of FBP on cameras A-1 and A-2, while no significant difference was detected on cameras B-1 and B-2. Good linearity of SBR was observed in all cameras. Inthe comparison between Chang�s AC and CTAC, a significant correlation was observed on all cameras. The difference in the background region influenced SBR differently in Chang�s AC and CTAC on cameras A-1 and B-1.Conclusion: Iterative reconstruction improved image quality on all cameras, although edge artifacts were observed in images captured by cameras with RR. The SBR of OSEM with RR was higher than that of FBP, while the SBR of OSEM without RR was equal to that of FBP. Also, the SBR of Chang�s AC varied with different background regions in cameras A-1 and B-1.

Item Type: Article
Subjects: WN Radiology . Diagnostic Imaging
Divisions: Journals > Asia Oceania J Nuclear Medicine & Biology
Depositing User: AOJNMB AOJNMB
Date Deposited: 19 Sep 2017 12:52
Last Modified: 19 Sep 2017 12:52
URI: http://eprints.mums.ac.ir/id/eprint/1010

Actions (login required)

View Item View Item