by Weiqiang Dou, Oliver Speck, Thomas Benner, Jörn Kaufmann, Meng Li, Kai Zhong, Martin Walter
Abstract:
OBJECT: The purpose of this study was to test, for the first time, whether spectroscopy voxels could be positioned automatically with high accuracy and reproducibility in ultrahigh-field longitudinal magnetic resonance spectroscopy (MRS) studies. MATERIALS AND METHODS: MRS voxels were automatically positioned in two cingulate subregions of 12 healthy subjects using a vendor-provided automatic voxel positioning (AutoAlign) technique, and were manually placed in the same regions of 10 healthy subjects by an experienced technician in three 7 T MRS scan sessions. Different coils were used for manual (24-channel coil) and automatic (32-channel coil) voxel placement, and the effects of signal-to-noise-ratio differences on the spectra were considered. RESULTS: Over three scan sessions and two regions scanned for each subject, a mean voxel geometric overlap ratio of 0.91 for automatic positioning reflected accurate voxel alignment, while the geometric overlap ratio was only 0.70 for voxels placed manually. Comparable voxel positions among the three scan sessions (p \textbackslashtextbackslashtextbackslashtextbackslashtextgreater 0.05) indicated high reproducibility of automatic voxel alignment. In comparison, significant voxel displacement among scan sessions (p \textbackslashtextbackslashtextbackslashtextbackslashtextless 0.05) was found using manual voxel positioning. CONCLUSIONS: In view of the highly accurate and reproducible voxel alignment with automatic voxel positioning, we propose the application of automatic rather than manual voxel positioning in future ultrahigh-field longitudinal MRS studies.
Reference:
Automatic voxel positioning for MRS at 7 T (Weiqiang Dou, Oliver Speck, Thomas Benner, Jörn Kaufmann, Meng Li, Kai Zhong, Martin Walter), In Magma (New York, N.Y.), volume 28, 2015.
Bibtex Entry:
@article{dou_automatic_2015,
	title = {Automatic voxel positioning for {MRS} at 7 {T}},
	volume = {28},
	issn = {1352-8661 0968-5243},
	doi = {10.1007/s10334-014-0469-9},
	abstract = {OBJECT: The purpose of this study was to test, for the first time, whether spectroscopy voxels could be positioned automatically with high accuracy and reproducibility in ultrahigh-field longitudinal magnetic resonance spectroscopy (MRS) studies. MATERIALS AND METHODS: MRS voxels were automatically positioned in two cingulate subregions of 12 healthy subjects using a vendor-provided automatic voxel positioning (AutoAlign) technique, and were manually placed in the same regions of 10 healthy subjects by an experienced technician in three 7 T MRS scan sessions. Different coils were used for manual (24-channel coil) and automatic (32-channel coil) voxel placement, and the effects of signal-to-noise-ratio differences on the spectra were considered. RESULTS: Over three scan sessions and two regions scanned for each subject, a mean voxel geometric overlap ratio of 0.91 for automatic positioning reflected accurate voxel alignment, while the geometric overlap ratio was only 0.70 for voxels placed manually. Comparable voxel positions among the three scan sessions (p {\textbackslash}textbackslashtextbackslashtextbackslashtextgreater 0.05) indicated high reproducibility of automatic voxel alignment. In comparison, significant voxel displacement among scan sessions (p {\textbackslash}textbackslashtextbackslashtextbackslashtextless 0.05) was found using manual voxel positioning. CONCLUSIONS: In view of the highly accurate and reproducible voxel alignment with automatic voxel positioning, we propose the application of automatic rather than manual voxel positioning in future ultrahigh-field longitudinal MRS studies.},
	language = {eng},
	number = {3},
	journal = {Magma (New York, N.Y.)},
	author = {Dou, Weiqiang and Speck, Oliver and Benner, Thomas and Kaufmann, Jörn and Li, Meng and Zhong, Kai and Walter, Martin},
	month = jun,
	year = {2015},
	pmid = {25408107},
	keywords = {*Subtraction Technique, Adult, Automated/methods, Female, Gyrus Cinguli/*anatomy \& histology/*chemistry, Humans, Image Enhancement/*methods, Imaging, Magnetic Resonance Imaging/methods, Magnetic Resonance Spectroscopy/*methods, Male, Molecular Imaging/methods, Pattern Recognition, Reproducibility of Results, Sensitivity and Specificity, Three-Dimensional/*methods},
	pages = {259--270}
}