HUSCAP logo Hokkaido Univ. logo

Hokkaido University Collection of Scholarly and Academic Papers >
Hokkaido University Hospital >
Peer-reviewed Journal Articles, etc >

Development of three-dimensional MR neurography using an optimized combination of compressed sensing and parallel imaging

This item is licensed under:Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International

Files in This Item:
Fig 1.pdf62.74 kBPDFView/Open
Fig 2(a) (b) (c).pdf166.9 kBPDFView/Open
Fig 3(a) (b) (c) (d) (e).pdf333.98 kBPDFView/Open
Fig 4(a) (b) (c).pdf285.35 kBPDFView/Open
Magn Reson Imaging 87 32-37.pdf290.22 kBPDFView/Open
Please use this identifier to cite or link to this item:

Title: Development of three-dimensional MR neurography using an optimized combination of compressed sensing and parallel imaging
Authors: Aoike, Takuya Browse this author
Fujima, Noriyuki Browse this author →KAKEN DB
Yoneyama, Masami Browse this author
Fujiwara, Taro Browse this author
Takamori, Sayaka Browse this author
Aoike, Suzuko Browse this author
Ishizaka, Kinya Browse this author
Kudo, Kohsuke Browse this author →KAKEN DB
Keywords: Magnetic resonance neurography
Compressed sensing
Sensitivity encoding
Cervical nerve
Issue Date: 1-Apr-2022
Publisher: Elsevier
Journal Title: Magnetic resonance imaging
Volume: 87
Start Page: 32
End Page: 37
Publisher DOI: 10.1016/j.mri.2021.12.002
Abstract: Purpose: To assess the cervical magnetic resonance neurography (MRN) imaging quality obtained with compressed sensing and sensitivity-encoding (compressed SENSE; CS-SENSE) technique in comparison to that obtained with the conventional parallel imaging (i.e., SENSE) technique. Materials and methods: Five healthy volunteers underwent a three-dimensional (3D) turbo spin-echo (TSE)-based cervical MRN examination using a 3.0 Tesla MR-unit. All MRN acquisitions were performed with CS-SENSE and conventional SENSE. We used four acceleration factors (4, 8, 16 and 32) in CS-SENSE. The image quality in MRN was evaluated by assessing the degree of cervical nerve depiction using the contrast ratio (CR) and contrast-noise ratio (CNR) between the cervical nerve and the background signal intensity and a visual scoring system (1: poor, 2: moderate, 3: good). In all of the CR, CNR and visual score, we calculated the ratio of the CS-SENSE-based MRN to that from SENSE-based MRN plus the 95% confidence intervals (CIs) of these ratios. Results: In the multiple comparison of MRN images with the control of conventional SENSE-based MRN, both the quantitative CR values and the visual score for the CS-SENSE factors of 16 and 32 were significantly lower, whereas the CS-SENSE factors of 4 and 8 showed a non-significant difference. In addition, the quantitative CNR values obtained with the CS-SENSE factors of 4 and 8 were significantly higher than that obtained with the conventional SENSE-based MRN while the CS-SENSE factor of 32 was significantly lower, in contrast, the CS SENSE factors of 16 showed a non-significant difference. For CS-SENSE factors of 4 and 8, all ratios of the CS SENSE-based MRN values for CR, CNR and visual scores to those from SENSE-based MRN were above 0.95. Conclusion: CS-SENSE-based MRN can accomplish fast scanning with sufficient image quality when using a high acceleration factor.
Rights: © 2022. This manuscript version is made available under the CC-BY-NC-ND 4.0 license
Type: article (author version)
Appears in Collections:北海道大学病院 (Hokkaido University Hospital) > 雑誌発表論文等 (Peer-reviewed Journal Articles, etc)

Submitter: 藤間 憲幸

Export metadata:

OAI-PMH ( junii2 , jpcoar_1.0 )

MathJax is now OFF:


 - Hokkaido University