Evaluation of Spatial Attentive Deep Learning for Automatic Placental Segmentation on Longitudinal MRI.
automated placental segmentation
convolutional neural network
spatial attentive deep learning
Journal
Journal of magnetic resonance imaging : JMRI
ISSN: 1522-2586
Titre abrégé: J Magn Reson Imaging
Pays: United States
ID NLM: 9105850
Informations de publication
Date de publication:
05 2023
05 2023
Historique:
revised:
03
08
2022
received:
13
05
2022
accepted:
05
08
2022
pmc-release:
01
05
2024
medline:
7
4
2023
entrez:
6
4
2023
pubmed:
7
4
2023
Statut:
ppublish
Résumé
Automated segmentation of the placenta by MRI in early pregnancy may help predict normal and aberrant placenta function, which could improve the efficiency of placental assessment and the prediction of pregnancy outcomes. An automated segmentation method that works at one gestational age may not transfer effectively to other gestational ages. To evaluate a spatial attentive deep learning method (SADL) for automated placental segmentation on longitudinal placental MRI scans. Prospective, single-center. A total of 154 pregnant women who underwent MRI scans at both 14-18 weeks of gestation and at 19-24 weeks of gestation, divided into training (N = 108), validation (N = 15), and independent testing datasets (N = 31). A 3 T, T2-weighted half Fourier single-shot turbo spin-echo (T2-HASTE) sequence. The reference standard of placental segmentation was manual delineation on T2-HASTE by a third-year neonatology clinical fellow (B.L.) under the supervision of an experienced maternal-fetal medicine specialist (C.J. with 20 years of experience) and an MRI scientist (K.S. with 19 years of experience). The three-dimensional Dice similarity coefficient (DSC) was used to measure the automated segmentation performance compared to the manual placental segmentation. A paired t-test was used to compare the DSCs between SADL and U-Net methods. A Bland-Altman plot was used to analyze the agreement between manual and automated placental volume measurements. A P value < 0.05 was considered statistically significant. In the testing dataset, SADL achieved average DSCs of 0.83 ± 0.06 and 0.84 ± 0.05 in the first and second MRI, which were significantly higher than those achieved by U-Net (0.77 ± 0.08 and 0.76 ± 0.10, respectively). A total of 6 out of 62 MRI scans (9.6%) had volume measurement differences between the SADL-based automated and manual volume measurements that were out of 95% limits of agreement. SADL can automatically detect and segment the placenta with high performance in MRI at two different gestational ages. 4 TECHNICAL EFFICACY STAGE: 2.
Sections du résumé
BACKGROUND
Automated segmentation of the placenta by MRI in early pregnancy may help predict normal and aberrant placenta function, which could improve the efficiency of placental assessment and the prediction of pregnancy outcomes. An automated segmentation method that works at one gestational age may not transfer effectively to other gestational ages.
PURPOSE
To evaluate a spatial attentive deep learning method (SADL) for automated placental segmentation on longitudinal placental MRI scans.
STUDY TYPE
Prospective, single-center.
SUBJECTS
A total of 154 pregnant women who underwent MRI scans at both 14-18 weeks of gestation and at 19-24 weeks of gestation, divided into training (N = 108), validation (N = 15), and independent testing datasets (N = 31).
FIELD STRENGTH/SEQUENCE
A 3 T, T2-weighted half Fourier single-shot turbo spin-echo (T2-HASTE) sequence.
ASSESSMENT
The reference standard of placental segmentation was manual delineation on T2-HASTE by a third-year neonatology clinical fellow (B.L.) under the supervision of an experienced maternal-fetal medicine specialist (C.J. with 20 years of experience) and an MRI scientist (K.S. with 19 years of experience).
STATISTICAL TESTS
The three-dimensional Dice similarity coefficient (DSC) was used to measure the automated segmentation performance compared to the manual placental segmentation. A paired t-test was used to compare the DSCs between SADL and U-Net methods. A Bland-Altman plot was used to analyze the agreement between manual and automated placental volume measurements. A P value < 0.05 was considered statistically significant.
RESULTS
In the testing dataset, SADL achieved average DSCs of 0.83 ± 0.06 and 0.84 ± 0.05 in the first and second MRI, which were significantly higher than those achieved by U-Net (0.77 ± 0.08 and 0.76 ± 0.10, respectively). A total of 6 out of 62 MRI scans (9.6%) had volume measurement differences between the SADL-based automated and manual volume measurements that were out of 95% limits of agreement.
DATA CONCLUSIONS
SADL can automatically detect and segment the placenta with high performance in MRI at two different gestational ages.
LEVEL OF EVIDENCE
4 TECHNICAL EFFICACY STAGE: 2.
Identifiants
pubmed: 37021577
doi: 10.1002/jmri.28403
pmc: PMC10080136
mid: NIHMS1830294
doi:
Types de publication
Journal Article
Research Support, N.I.H., Extramural
Langues
eng
Sous-ensembles de citation
IM
Pagination
1533-1540Subventions
Organisme : NICHD NIH HHS
ID : U01 HD087221
Pays : United States
Commentaires et corrections
Type : CommentIn
Informations de copyright
© 2022 International Society for Magnetic Resonance in Medicine.
Références
Ultrasound Obstet Gynecol. 2020 Mar;55(3):293-302
pubmed: 31452271
IEEE Access. 2020;8:151817-151828
pubmed: 33564563
Med Image Anal. 2016 Dec;34:137-147
pubmed: 27179367
Proc SPIE Int Soc Opt Eng. 2020 Feb;11314:
pubmed: 32476702
Cancer Res. 2017 Nov 1;77(21):e104-e107
pubmed: 29092951
Med Image Anal. 2017 Jan;35:18-31
pubmed: 27310171
JCI Insight. 2018 Jun 7;3(11):
pubmed: 29875312
Sci Rep. 2017 Jul 11;7(1):5110
pubmed: 28698556
IEEE Trans Pattern Anal Mach Intell. 2019 Jul;41(7):1559-1572
pubmed: 29993532
IEEE Trans Pattern Anal Mach Intell. 2020 Jul 03;PP:
pubmed: 32750802
J Magn Reson Imaging. 2018 Feb;47(2):449-458
pubmed: 28734056
AJNR Am J Neuroradiol. 1997 Oct;18(9):1635-40
pubmed: 9367310
Semin Perinatol. 2014 Apr;38(3):146-50
pubmed: 24836826
IEEE Trans Med Imaging. 2018 Jul;37(7):1562-1573
pubmed: 29969407
Hum Brain Mapp. 2014 Aug;35(8):4236-48
pubmed: 24532539
J Magn Reson Imaging. 2020 Apr;51(4):1247-1257
pubmed: 31680405
Trends Biochem Sci. 1990 Dec;15(12):455-8
pubmed: 2077683
Annu Int Conf IEEE Eng Med Biol Soc. 2019 Jul;2019:6718-6723
pubmed: 31947383
Magn Reson Imaging. 1999 Sep;17(7):965-71
pubmed: 10463645
Biochem Med (Zagreb). 2015 Jun 05;25(2):141-51
pubmed: 26110027
Placenta. 2001 Jul;22(6):602-5
pubmed: 11440550