| Literature DB >> 30797888 |
Qiqi Tong1, Hongjian He2, Ting Gong3, Chen Li4, Peipeng Liang5, Tianyi Qian6, Yi Sun7, Qiuping Ding8, Kuncheng Li9, Jianhui Zhong10.
Abstract
Reproducibility of multicenter diffusion magnetic resonance imaging has drawn more attention recently due to rapidly increasing need for large-size brain imaging studies. Advanced multi-shell diffusion models are recommended for their potentials to provide variety of physio-pathological information. While previous studies have investigated the consistency of single-shell diffusion acquisition from various hardware and protocols, a well-controlled study with multi-shell acquisition would be necessary to understand the inherent factors of reproducibility from new complexity of such acquisition protocol. In this study, three traveling subjects were scanned at eight imaging centers equipped with the same type of scanners using the same multi-shell diffusion imaging protocol. Track density imaging and structure connectomes were investigated in local-scale distribution and in distal-scale connectivity, respectively. With evaluations of the coefficient of variation and the intra-class correlation coefficient, our results indicated: 1) similar to single-shell schemes, the intra-center reproducibility of multi-shell is higher than inter-center; 2) multi-shell schemes produce higher reproducibility and precision among centers compared to the single-shell schemes; and 3) in addition to the diffusion schemes, image quality and the presence of complex fiber structure could also associated with multicenter reproducibility.Keywords: Diffusion MRI; Multi-center studies; Multi-shell diffusion; Reproducibility; Structural connectome; Track-density imaging
Mesh:
Year: 2019 PMID: 30797888 DOI: 10.1016/j.mri.2019.02.011
Source DB: PubMed Journal: Magn Reson Imaging ISSN: 0730-725X Impact factor: 2.546