Modelling brain development to detect white matter injury in term and preterm born neonates

Jonathan O'Muircheartaigh; Emma Robinson; Maximilian Pietsch; Thomas Wolfers; Paul Aljabar; Lucilio Cordero Grande; Rui Gomes Teixeira; Jelena Bozek; Andreas Schuh; Antonios Makropoulos; Dafnis Batalle; Katy Vecchiato; Johannes K. Steinweg; Sean Fitzgibbon; Emer Hughes; Anthony Price; Andre Marquand; Daniel  Rueckert; Mary Rutherford; Jo Hajnal; Serena J Counsell; David Edwards

doi:10.1093/brain/awz412

Modelling brain development to detect white matter injury in term and preterm born neonates

Jonathan O'Muircheartaigh, Emma Robinson, Maximilian Pietsch, Thomas Wolfers, Paul Aljabar, Lucilio Cordero Grande, Rui Gomes Teixeira, Jelena Bozek, Andreas Schuh, Antonios Makropoulos, Dafnis Batalle, Katy Vecchiato, Johannes K. Steinweg, Sean Fitzgibbon, Emer Hughes, Anthony Price, Andre Marquand, Daniel Rueckert, Mary Rutherford, Jo HajnalSerena J Counsell, David Edwards

Research output: Contribution to journal › Article › peer-review

35 Citations (Scopus)

192 Downloads (Pure)

Abstract

Premature birth occurs during a period of rapid brain growth. In this context, interpreting clinical neuroimaging can be complicated by the typical changes in brain contrast, size and gyrification occurring in the background to any pathology. To model and describe this evolving background in brain shape and contrast, we used a Bayesian regression technique, Gaussian process regression, adapted to multiple correlated outputs. Using MRI, we simultaneously estimated brain tissue intensity on T ₁- and T ₂weighted scans as well as local tissue shape in a large cohort of 408 neonates scanned cross-sectionally across the perinatal period. The resulting model provided a continuous estimate of brain shape and intensity, appropriate to age at scan, degree of prematurity and sex. Next, we investigated the clinical utility of this model to detect focal white matter injury. In individual neonates, we calculated deviations of a neonate’s observed MRI from that predicted by the model to detect punctate white matter lesions with very good accuracy (area under the curve 4 0.95). To investigate longitudinal consistency of the model, we calculated model deviations in 46 neonates who were scanned on a second occasion. These infants’ voxelwise deviations from the model could be used to identify them from the other 408 images in 83% (T ₂-weighted) and 76% (T ₁-weighted) of cases, indicating an anatomical fingerprint. Our approach provides accurate estimates of non-linear changes in brain tissue intensity and shape with clear potential for radiological use.

Original language	English
Pages (from-to)	467-479
Number of pages	13
Journal	Brain : a journal of neurology
Volume	143
Issue number	2
Early online date	16 Jan 2020
DOIs	https://doi.org/10.1093/brain/awz412
Publication status	Published - 1 Feb 2020

Keywords

brain development
imaging methodology
neonatology
neuroanatomy
neuropathology

Access to Document

10.1093/brain/awz412Licence: CC BY

Modelling brain development to_O'MUIRCHEARTAIGH_Acc19Nov2019Epub16Jan2020_GOLD VoR(CC BY)Final published version, 1.23 MBLicence: CC BY

Cite this

O'Muircheartaigh, J., Robinson, E., Pietsch, M., Wolfers, T., Aljabar, P., Cordero Grande, L., Gomes Teixeira, R., Bozek, J., Schuh, A., Makropoulos, A., Batalle, D., Vecchiato, K., Steinweg, J. K., Fitzgibbon, S., Hughes, E., Price, A., Marquand, A., Rueckert, D., Rutherford, M., ... Edwards, D. (2020). Modelling brain development to detect white matter injury in term and preterm born neonates. Brain : a journal of neurology, 143(2), 467-479. https://doi.org/10.1093/brain/awz412

@article{6ab7496631e441a7b1110514d890bca9,

title = "Modelling brain development to detect white matter injury in term and preterm born neonates",

abstract = "Premature birth occurs during a period of rapid brain growth. In this context, interpreting clinical neuroimaging can be complicated by the typical changes in brain contrast, size and gyrification occurring in the background to any pathology. To model and describe this evolving background in brain shape and contrast, we used a Bayesian regression technique, Gaussian process regression, adapted to multiple correlated outputs. Using MRI, we simultaneously estimated brain tissue intensity on T 1- and T 2weighted scans as well as local tissue shape in a large cohort of 408 neonates scanned cross-sectionally across the perinatal period. The resulting model provided a continuous estimate of brain shape and intensity, appropriate to age at scan, degree of prematurity and sex. Next, we investigated the clinical utility of this model to detect focal white matter injury. In individual neonates, we calculated deviations of a neonate{\textquoteright}s observed MRI from that predicted by the model to detect punctate white matter lesions with very good accuracy (area under the curve 4 0.95). To investigate longitudinal consistency of the model, we calculated model deviations in 46 neonates who were scanned on a second occasion. These infants{\textquoteright} voxelwise deviations from the model could be used to identify them from the other 408 images in 83% (T 2-weighted) and 76% (T 1-weighted) of cases, indicating an anatomical fingerprint. Our approach provides accurate estimates of non-linear changes in brain tissue intensity and shape with clear potential for radiological use. ",

keywords = "brain development, imaging methodology, neonatology, neuroanatomy, neuropathology",

author = "Jonathan O'Muircheartaigh and Emma Robinson and Maximilian Pietsch and Thomas Wolfers and Paul Aljabar and {Cordero Grande}, Lucilio and {Gomes Teixeira}, Rui and Jelena Bozek and Andreas Schuh and Antonios Makropoulos and Dafnis Batalle and Katy Vecchiato and Steinweg, {Johannes K.} and Sean Fitzgibbon and Emer Hughes and Anthony Price and Andre Marquand and Daniel Rueckert and Mary Rutherford and Jo Hajnal and Counsell, {Serena J} and David Edwards",

year = "2020",

month = feb,

day = "1",

doi = "10.1093/brain/awz412",

language = "English",

volume = "143",

pages = "467--479",

journal = "Brain : a journal of neurology",

issn = "0006-8950",

publisher = "Oxford University Press (OUP)",

number = "2",

}

O'Muircheartaigh, J , Robinson, E , Pietsch, M, Wolfers, T, Aljabar, P , Cordero Grande, L , Gomes Teixeira, R, Bozek, J, Schuh, A, Makropoulos, A, Batalle, D , Vecchiato, K , Steinweg, JK, Fitzgibbon, S, Hughes, E, Price, A , Marquand, A, Rueckert, D, Rutherford, M , Hajnal, J , Counsell, SJ & Edwards, D 2020, 'Modelling brain development to detect white matter injury in term and preterm born neonates', Brain : a journal of neurology, vol. 143, no. 2, pp. 467-479. https://doi.org/10.1093/brain/awz412

TY - JOUR

T1 - Modelling brain development to detect white matter injury in term and preterm born neonates

AU - O'Muircheartaigh, Jonathan

AU - Robinson, Emma

AU - Pietsch, Maximilian

AU - Wolfers, Thomas

AU - Aljabar, Paul

AU - Cordero Grande, Lucilio

AU - Gomes Teixeira, Rui

AU - Bozek, Jelena

AU - Schuh, Andreas

AU - Makropoulos, Antonios

AU - Batalle, Dafnis

AU - Vecchiato, Katy

AU - Steinweg, Johannes K.

AU - Fitzgibbon, Sean

AU - Hughes, Emer

AU - Price, Anthony

AU - Marquand, Andre

AU - Rueckert, Daniel

AU - Rutherford, Mary

AU - Hajnal, Jo

AU - Counsell, Serena J

AU - Edwards, David

PY - 2020/2/1

Y1 - 2020/2/1

N2 - Premature birth occurs during a period of rapid brain growth. In this context, interpreting clinical neuroimaging can be complicated by the typical changes in brain contrast, size and gyrification occurring in the background to any pathology. To model and describe this evolving background in brain shape and contrast, we used a Bayesian regression technique, Gaussian process regression, adapted to multiple correlated outputs. Using MRI, we simultaneously estimated brain tissue intensity on T 1- and T 2weighted scans as well as local tissue shape in a large cohort of 408 neonates scanned cross-sectionally across the perinatal period. The resulting model provided a continuous estimate of brain shape and intensity, appropriate to age at scan, degree of prematurity and sex. Next, we investigated the clinical utility of this model to detect focal white matter injury. In individual neonates, we calculated deviations of a neonate’s observed MRI from that predicted by the model to detect punctate white matter lesions with very good accuracy (area under the curve 4 0.95). To investigate longitudinal consistency of the model, we calculated model deviations in 46 neonates who were scanned on a second occasion. These infants’ voxelwise deviations from the model could be used to identify them from the other 408 images in 83% (T 2-weighted) and 76% (T 1-weighted) of cases, indicating an anatomical fingerprint. Our approach provides accurate estimates of non-linear changes in brain tissue intensity and shape with clear potential for radiological use.

AB - Premature birth occurs during a period of rapid brain growth. In this context, interpreting clinical neuroimaging can be complicated by the typical changes in brain contrast, size and gyrification occurring in the background to any pathology. To model and describe this evolving background in brain shape and contrast, we used a Bayesian regression technique, Gaussian process regression, adapted to multiple correlated outputs. Using MRI, we simultaneously estimated brain tissue intensity on T 1- and T 2weighted scans as well as local tissue shape in a large cohort of 408 neonates scanned cross-sectionally across the perinatal period. The resulting model provided a continuous estimate of brain shape and intensity, appropriate to age at scan, degree of prematurity and sex. Next, we investigated the clinical utility of this model to detect focal white matter injury. In individual neonates, we calculated deviations of a neonate’s observed MRI from that predicted by the model to detect punctate white matter lesions with very good accuracy (area under the curve 4 0.95). To investigate longitudinal consistency of the model, we calculated model deviations in 46 neonates who were scanned on a second occasion. These infants’ voxelwise deviations from the model could be used to identify them from the other 408 images in 83% (T 2-weighted) and 76% (T 1-weighted) of cases, indicating an anatomical fingerprint. Our approach provides accurate estimates of non-linear changes in brain tissue intensity and shape with clear potential for radiological use.

KW - brain development

KW - imaging methodology

KW - neonatology

KW - neuroanatomy

KW - neuropathology

UR - http://www.scopus.com/inward/record.url?scp=85079247955&partnerID=8YFLogxK

U2 - 10.1093/brain/awz412

DO - 10.1093/brain/awz412

M3 - Article

SN - 0006-8950

VL - 143

SP - 467

EP - 479

JO - Brain : a journal of neurology

JF - Brain : a journal of neurology

IS - 2

ER -

Modelling brain development to detect white matter injury in term and preterm born neonates

Abstract

Keywords

Access to Document

Other files and links

Fingerprint

Cite this