Applicability and reproducibility of 2D multi-slice GRASE myelin water fraction with varying acquisition acceleration

Gerhard S. Drenthen, Walter H. Backes, Albert P. Aldenkamp, Jacobus F.A. Jansen (Corresponding author)

Research output: Contribution to journalArticleAcademicpeer-review

1 Citation (Scopus)

Abstract

Non-invasive quantification of the in vivo myelin content may provide valuable information regarding healthy maturation of the brain, as well as insights into demyelination of several neurological disorders. However, these scans are often long thereby limiting acquisition of large brain parts in clinically feasible acquisition times. Therefore, fast acquisition of whole brain myelin content is important. To avoid errors related to slice-selective pulses, most of the previous whole brain studies on myelin content relied on a 3D acquisition. However, multi-slice (2D) acquisition methods are often faster, and less susceptible to motion artifacts. Therefore, multi-slice approaches can be beneficial in a clinical setting. We investigated the applicability and reproducibility of whole brain multi-slice GRASE myelin-water imaging with post-acquisition slice-profile correction in healthy volunteers (aged 25-32y). The applicability was evaluated using the agreement between the multi-slice GRASE and the reference method for myelin-water imaging, single-slice multi spin-echo (MSE) acquisition. Additionally, we assessed the effect of varying acquisition acceleration using parallel imaging on the reproducibility values. First, the multi-slice myelin-water maps showed good agreement with the single-slice reference method, with a bias of at most 1.2% in absolute MWF values. Second, we found an average within-subject coefficient of variation (CoV) of 5.9% and an average intra-class correlation coefficient (ICC) of 0.90 for myelin-water estimation using a multi-slice GRASE sequence without parallel acceleration (scan time 14:06 min), while acquisition with a parallel acceleration factor of 2 resulted in a slightly worse average within-subject CoV of 6.4% and an average ICC of 0.83 at half the scan time. Hence, a multi-slice GRASE acquisition with parallel acceleration factor 2 and a scan time of 7:30 min still provides an excellent reproducibility.

LanguageEnglish
Pages333-339
Number of pages7
JournalNeuroimage
Volume195
DOIs
StatePublished - 15 Jul 2019

Fingerprint

Myelin Sheath
Water
Brain
Demyelinating Diseases
Nervous System Diseases
Artifacts
Healthy Volunteers

Keywords

  • Myelin water fraction
  • Reproducibility
  • T2 relaxometry

Cite this

@article{4ce227ce1011466aaaa5c76afd6a242b,
title = "Applicability and reproducibility of 2D multi-slice GRASE myelin water fraction with varying acquisition acceleration",
abstract = "Non-invasive quantification of the in vivo myelin content may provide valuable information regarding healthy maturation of the brain, as well as insights into demyelination of several neurological disorders. However, these scans are often long thereby limiting acquisition of large brain parts in clinically feasible acquisition times. Therefore, fast acquisition of whole brain myelin content is important. To avoid errors related to slice-selective pulses, most of the previous whole brain studies on myelin content relied on a 3D acquisition. However, multi-slice (2D) acquisition methods are often faster, and less susceptible to motion artifacts. Therefore, multi-slice approaches can be beneficial in a clinical setting. We investigated the applicability and reproducibility of whole brain multi-slice GRASE myelin-water imaging with post-acquisition slice-profile correction in healthy volunteers (aged 25-32y). The applicability was evaluated using the agreement between the multi-slice GRASE and the reference method for myelin-water imaging, single-slice multi spin-echo (MSE) acquisition. Additionally, we assessed the effect of varying acquisition acceleration using parallel imaging on the reproducibility values. First, the multi-slice myelin-water maps showed good agreement with the single-slice reference method, with a bias of at most 1.2{\%} in absolute MWF values. Second, we found an average within-subject coefficient of variation (CoV) of 5.9{\%} and an average intra-class correlation coefficient (ICC) of 0.90 for myelin-water estimation using a multi-slice GRASE sequence without parallel acceleration (scan time 14:06 min), while acquisition with a parallel acceleration factor of 2 resulted in a slightly worse average within-subject CoV of 6.4{\%} and an average ICC of 0.83 at half the scan time. Hence, a multi-slice GRASE acquisition with parallel acceleration factor 2 and a scan time of 7:30 min still provides an excellent reproducibility.",
keywords = "Myelin water fraction, Reproducibility, T2 relaxometry",
author = "Drenthen, {Gerhard S.} and Backes, {Walter H.} and Aldenkamp, {Albert P.} and Jansen, {Jacobus F.A.}",
year = "2019",
month = "7",
day = "15",
doi = "10.1016/j.neuroimage.2019.04.011",
language = "English",
volume = "195",
pages = "333--339",
journal = "Neuroimage",
issn = "1053-8119",
publisher = "Academic Press Inc.",

}

Applicability and reproducibility of 2D multi-slice GRASE myelin water fraction with varying acquisition acceleration. / Drenthen, Gerhard S.; Backes, Walter H.; Aldenkamp, Albert P.; Jansen, Jacobus F.A. (Corresponding author).

In: Neuroimage, Vol. 195, 15.07.2019, p. 333-339.

Research output: Contribution to journalArticleAcademicpeer-review

TY - JOUR

T1 - Applicability and reproducibility of 2D multi-slice GRASE myelin water fraction with varying acquisition acceleration

AU - Drenthen,Gerhard S.

AU - Backes,Walter H.

AU - Aldenkamp,Albert P.

AU - Jansen,Jacobus F.A.

PY - 2019/7/15

Y1 - 2019/7/15

N2 - Non-invasive quantification of the in vivo myelin content may provide valuable information regarding healthy maturation of the brain, as well as insights into demyelination of several neurological disorders. However, these scans are often long thereby limiting acquisition of large brain parts in clinically feasible acquisition times. Therefore, fast acquisition of whole brain myelin content is important. To avoid errors related to slice-selective pulses, most of the previous whole brain studies on myelin content relied on a 3D acquisition. However, multi-slice (2D) acquisition methods are often faster, and less susceptible to motion artifacts. Therefore, multi-slice approaches can be beneficial in a clinical setting. We investigated the applicability and reproducibility of whole brain multi-slice GRASE myelin-water imaging with post-acquisition slice-profile correction in healthy volunteers (aged 25-32y). The applicability was evaluated using the agreement between the multi-slice GRASE and the reference method for myelin-water imaging, single-slice multi spin-echo (MSE) acquisition. Additionally, we assessed the effect of varying acquisition acceleration using parallel imaging on the reproducibility values. First, the multi-slice myelin-water maps showed good agreement with the single-slice reference method, with a bias of at most 1.2% in absolute MWF values. Second, we found an average within-subject coefficient of variation (CoV) of 5.9% and an average intra-class correlation coefficient (ICC) of 0.90 for myelin-water estimation using a multi-slice GRASE sequence without parallel acceleration (scan time 14:06 min), while acquisition with a parallel acceleration factor of 2 resulted in a slightly worse average within-subject CoV of 6.4% and an average ICC of 0.83 at half the scan time. Hence, a multi-slice GRASE acquisition with parallel acceleration factor 2 and a scan time of 7:30 min still provides an excellent reproducibility.

AB - Non-invasive quantification of the in vivo myelin content may provide valuable information regarding healthy maturation of the brain, as well as insights into demyelination of several neurological disorders. However, these scans are often long thereby limiting acquisition of large brain parts in clinically feasible acquisition times. Therefore, fast acquisition of whole brain myelin content is important. To avoid errors related to slice-selective pulses, most of the previous whole brain studies on myelin content relied on a 3D acquisition. However, multi-slice (2D) acquisition methods are often faster, and less susceptible to motion artifacts. Therefore, multi-slice approaches can be beneficial in a clinical setting. We investigated the applicability and reproducibility of whole brain multi-slice GRASE myelin-water imaging with post-acquisition slice-profile correction in healthy volunteers (aged 25-32y). The applicability was evaluated using the agreement between the multi-slice GRASE and the reference method for myelin-water imaging, single-slice multi spin-echo (MSE) acquisition. Additionally, we assessed the effect of varying acquisition acceleration using parallel imaging on the reproducibility values. First, the multi-slice myelin-water maps showed good agreement with the single-slice reference method, with a bias of at most 1.2% in absolute MWF values. Second, we found an average within-subject coefficient of variation (CoV) of 5.9% and an average intra-class correlation coefficient (ICC) of 0.90 for myelin-water estimation using a multi-slice GRASE sequence without parallel acceleration (scan time 14:06 min), while acquisition with a parallel acceleration factor of 2 resulted in a slightly worse average within-subject CoV of 6.4% and an average ICC of 0.83 at half the scan time. Hence, a multi-slice GRASE acquisition with parallel acceleration factor 2 and a scan time of 7:30 min still provides an excellent reproducibility.

KW - Myelin water fraction

KW - Reproducibility

KW - T2 relaxometry

UR - http://www.scopus.com/inward/record.url?scp=85064039170&partnerID=8YFLogxK

U2 - 10.1016/j.neuroimage.2019.04.011

DO - 10.1016/j.neuroimage.2019.04.011

M3 - Article

VL - 195

SP - 333

EP - 339

JO - Neuroimage

T2 - Neuroimage

JF - Neuroimage

SN - 1053-8119

ER -