Improved neonatal brain MRI segmentation by interpolation of motion corrupted slices.
MRI
interpolation
motion artifacts
neonatal
segmentation
Journal
Journal of neuroimaging : official journal of the American Society of Neuroimaging
ISSN: 1552-6569
Titre abrégé: J Neuroimaging
Pays: United States
ID NLM: 9102705
Informations de publication
Date de publication:
05 2022
05 2022
Historique:
revised:
21
02
2022
received:
22
06
2021
accepted:
22
02
2022
pubmed:
8
3
2022
medline:
12
5
2022
entrez:
7
3
2022
Statut:
ppublish
Résumé
To apply and evaluate an intensity-based interpolation technique, enabling segmentation of motion-affected neonatal brain MRI. Moderate-late preterm infants were enrolled in a prospective cohort study (Brain Imaging in Moderate-late Preterm infants "BIMP-study") between August 2017 and November 2019. T2-weighted MRI was performed around term equivalent age on a 3T MRI. Scans without motion (n = 27 [24%], control group) and with moderate-severe motion (n = 33 [29%]) were included. Motion-affected slices were re-estimated using intensity-based shape-preserving cubic spline interpolation, and automatically segmented in eight structures. Quality of interpolation and segmentation was visually assessed for errors after interpolation. Reliability was tested using interpolated control group scans (18/54 axial slices). Structural similarity index (SSIM) was used to compare T2-weighted scans, and Sørensen-Dice was used to compare segmentation before and after interpolation. Finally, volumes of brain structures of the control group were used assessing sensitivity (absolute mean fraction difference) and bias (confidence interval of mean difference). Visually, segmentation of 25 scans (22%) with motion artifacts improved with interpolation, while segmentation of eight scans (7%) with adjacent motion-affected slices did not improve. Average SSIM was .895 and Sørensen-Dice coefficients ranged between .87 and .97. Absolute mean fraction difference was ≤0.17 for less than or equal to five interpolated slices. Confidence intervals revealed a small bias for cortical gray matter (0.14-3.07 cm According to qualitative and quantitative assessment, intensity-based interpolation reduced the percentage of discarded scans from 29% to 7%.
Sections du résumé
BACKGROUND AND PURPOSE
To apply and evaluate an intensity-based interpolation technique, enabling segmentation of motion-affected neonatal brain MRI.
METHODS
Moderate-late preterm infants were enrolled in a prospective cohort study (Brain Imaging in Moderate-late Preterm infants "BIMP-study") between August 2017 and November 2019. T2-weighted MRI was performed around term equivalent age on a 3T MRI. Scans without motion (n = 27 [24%], control group) and with moderate-severe motion (n = 33 [29%]) were included. Motion-affected slices were re-estimated using intensity-based shape-preserving cubic spline interpolation, and automatically segmented in eight structures. Quality of interpolation and segmentation was visually assessed for errors after interpolation. Reliability was tested using interpolated control group scans (18/54 axial slices). Structural similarity index (SSIM) was used to compare T2-weighted scans, and Sørensen-Dice was used to compare segmentation before and after interpolation. Finally, volumes of brain structures of the control group were used assessing sensitivity (absolute mean fraction difference) and bias (confidence interval of mean difference).
RESULTS
Visually, segmentation of 25 scans (22%) with motion artifacts improved with interpolation, while segmentation of eight scans (7%) with adjacent motion-affected slices did not improve. Average SSIM was .895 and Sørensen-Dice coefficients ranged between .87 and .97. Absolute mean fraction difference was ≤0.17 for less than or equal to five interpolated slices. Confidence intervals revealed a small bias for cortical gray matter (0.14-3.07 cm
CONCLUSION
According to qualitative and quantitative assessment, intensity-based interpolation reduced the percentage of discarded scans from 29% to 7%.
Identifiants
pubmed: 35253956
doi: 10.1111/jon.12985
pmc: PMC9314603
doi:
Types de publication
Journal Article
Research Support, Non-U.S. Gov't
Langues
eng
Sous-ensembles de citation
IM
Pagination
480-492Informations de copyright
© 2022 The Authors. Journal of Neuroimaging published by Wiley Periodicals LLC on behalf of American Society of Neuroimaging.
Références
Clin Anat. 2015 Mar;28(2):168-83
pubmed: 25043926
Pediatr Radiol. 2012 Oct;42(10):1183-94
pubmed: 22886375
Clin Pediatr (Phila). 2017 Oct;56(12):1095-1103
pubmed: 27872353
IEEE Trans Med Imaging. 1996;15(6):881-92
pubmed: 18215967
Neuropsychol Rev. 2010 Dec;20(4):327-48
pubmed: 21042938
Front Neuroinform. 2016 Mar 29;10:12
pubmed: 27065840
Magn Reson Med. 2019 Oct;82(4):1527-1540
pubmed: 31081955
Magn Reson Med. 2018 Mar;79(3):1365-1376
pubmed: 28626962
Magn Reson Imaging. 2011 Feb;29(2):230-42
pubmed: 21129881
Proc SPIE Int Soc Opt Eng. 2010 Mar 11;7628:
pubmed: 24353379
Phys Med Biol. 2016 Mar 7;61(5):R32-56
pubmed: 26864183
IEEE Trans Med Imaging. 1996;15(4):580-7
pubmed: 18215939
Eur J Radiol. 2021 Mar;136:109500
pubmed: 33429207
IEEE Trans Med Imaging. 1999 Nov;18(11):1049-75
pubmed: 10661324
IEEE Trans Image Process. 2004 Apr;13(4):600-12
pubmed: 15376593
Comput Med Imaging Graph. 2012 Mar;36(2):130-8
pubmed: 21920702
J Magn Reson Imaging. 2015 Oct;42(4):887-901
pubmed: 25630632
Magn Reson Med. 2022 Jan;87(1):163-178
pubmed: 34390505
IEEE Trans Med Imaging. 2002 Dec;21(12):1479-89
pubmed: 12588032
AJNR Am J Neuroradiol. 2020 Mar;41(3):416-423
pubmed: 32054615
Med Image Anal. 2012 Dec;16(8):1550-64
pubmed: 22939612
Magn Reson Med. 1999 Nov;42(5):963-9
pubmed: 10542356
IEEE Trans Med Imaging. 2000 Jul;19(7):711-21
pubmed: 11055786
Magn Reson Med. 2020 Aug;84(2):
pubmed: 31898832
Eur J Radiol. 2008 Nov;68(2):309-19
pubmed: 18768276
Pediatr Radiol. 2008 Mar;38(3):260-4
pubmed: 18175110
Neuroimage. 2005 Jul 1;26(3):839-51
pubmed: 15955494
J Neuroimaging. 2022 May;32(3):480-492
pubmed: 35253956
IEEE Trans Med Imaging. 1992;11(4):507-16
pubmed: 18222892