Comparison between MRI-based attenuation correction methods for brain PET in dementia patients
The combination of Positron Emission Tomography (PET) with magnetic resonance imaging (MRI) in hybrid PET/MRI scanners offers a number of advantages in investigating brain structure and function. A critical step of PET data reconstruction is attenuation correction (AC). Accounting for bone in attenuation maps (μ-map) was shown to be important in brain PET studies. While there are a number of MRI-based AC methods, no systematic comparison between them has been performed so far. The aim of this work was to study the different performance obtained by some of the recent methods presented in the literature. To perform such a comparison, we focused on [18F]-Fluorodeoxyglucose-PET/MRI neurodegenerative dementing disorders, which are known to exhibit reduced levels of glucose metabolism in certain brain regions.
Four novel methods were used to calculate μ-maps from MRI data of 15 patients with Alzheimer’s dementia (AD). The methods cover two atlas-based methods, a segmentation method, and a hybrid template/segmentation method. Additionally, the Dixon-based and a UTE-based method, offered by a vendor, were included in the comparison. Performance was assessed at three levels: tissue identification accuracy in the μ-map, quantitative accuracy of reconstructed PET data in specific brain regions, and precision in diagnostic images at identifying hypometabolic areas.
Quantitative regional errors of −20–−10 % were obtained using the vendor’s AC methods, whereas the novel methods produced errors in a margin of ±5 %. The obtained precision at identifying areas with abnormally low levels of glucose uptake, potentially regions affected by AD, were 62.9 and 79.5 % for the two vendor AC methods, the former ignoring bone and the latter including bone information. The precision increased to 87.5–93.3 % in average for the four new methods, exhibiting similar performances.
We confirm that the AC methods based on the Dixon and UTE sequences provided by the vendor are inferior to alternative techniques. As a novel finding, there was no substantial difference between the recently proposed atlas-based, template-based and segmentation-based methods.
KeywordsAttenuation correction PET/MRI FDG-PET Neurostat Alzheimer’s dementia
The PET/MRI facility at the Technische Universität München was funded by the Großgeräteinitiative from the Deutsche Forschungsgemeinschaft, (DFG). The research leading to these results has received funding from the European Union Seventh Framework Program (FP7) under Grant Agreement n° 602621- Trimage, n° 294582- MUMI, and from the DFG grant no. FO 886/1–1. We thank David Izquierdo and Ninon Burgos for their assistance using their algorithms. We also thank Sylvia Schachoff and Claudia Meisinger for their technical assistance with the PET/MRI scanner.
Compliance with ethical standards
The research leading to these results has received funding from the Seventh Framework Programme [FP7/2007-2013] under grant agreement n° 602621- Trimage, n° 294582- MUMI, and from the DFG grant no. FO 886/1–1.
Conflict of interest
The authors declare that they have no conflict of interest.
All procedures performed in studies involving human participants were in accordance with the ethical standards of the institutional and/or national research committee and with the 1964 Helsinki declaration and its later amendments or comparable ethical standards.
Informed consent was obtained from all individual participants included in the study.
- 20.Delso G, Zeimpekis K, Carl F, Wiesinger F, Hüllner M, Veit-Haibach P. Cluster-based segmentation of dual-echo ultra-short echo time images for PET/MR bone localization. Eur J Nucl Med Mol Imaging. 2014;1:7.Google Scholar
- 40.Ladefoged CN, Hansen AE, Kelelr SH, et al. Dental artifacts in the head and neck region: implications for Dixon-based attenuation correction in PET/MR. Eur J Nucl Med Mol Imaging. 2015;2:8.Google Scholar