NeuroImage xxx (2010) xxx–xxx

YNIMG-07282; No. of pages: 6; 4C: 3

Contents lists available at ScienceDirect

NeuroImage

j ourna l homepage: www.e lsev ie r.com/ locate /yn img

ARTICLE IN PRESS

A common neonatal image phenotype predicts adverse neurodevelopmentaloutcome in children born preterm

J.P. Boardman a,d,⁎, C. Craven a, S. Valappil b, S.J. Counsell a, L.E. Dyet a, D. Rueckert c, P. Aljabar c,M.A. Rutherford a, A.T.M. Chew a,b, J.M. Allsop a, F. Cowan a,b, A.D. Edwards a,b

a Institute of Clinical Sciences, Imperial College London and MRC Clinical Sciences Centre, Hammersmith Hospital, Du Cane Road, London W12 0NN, UKb Division of Neonatology, Imperial College Healthcare NHS Trust, Hammersmith Hospital, Du Cane Road, London W12 0NN, UKc Department of Computing, Imperial College London, 180 Queen's Gate, London SW7 2AZ, UKd Simpson Centre for Reproductive Health, Royal Infirmary of Edinburgh, Edinburgh EH16 4SA, UK

⁎ Corresponding author. Department of Neonatology,tive Health, Royal Infirmary of Edinburgh, 51 Little Fra4SA, UK. Fax: +44 131 242 2674.

E-mail address: j.boardman@imperial.ac.uk (J.P. Boa

1053-8119/$ – see front matter © 2010 Elsevier Inc. Aldoi:10.1016/j.neuroimage.2010.04.261

Please cite this article as: Boardman, J.P., echildren born preterm, NeuroImage (2010

a b s t r a c t

a r t i c l e i n f o

Article history:Received 3 December 2009Revised 5 April 2010Accepted 28 April 2010Available online xxxx

Diffuse white matter injury is common in preterm infants and is a candidate substrate for later cognitiveimpairment. This injury pattern is associated with morphological changes in deep grey nuclei, thelocalization of which is uncertain. We test the hypotheses that diffuse white matter injury is associated withdiscrete focal tissue loss, and that this image phenotype is associated with impairment at 2 years.We acquired magnetic resonance images from 80 preterm infants at term equivalent (mean gestational age29+6 weeks) and 20 control infants (mean GA 39+2 weeks). Diffuse white matter injury was defined byabnormal apparent diffusion coefficient values in one or more white matter region (frontal, central orposterior white matter at the level of the centrum semiovale), and morphological difference between groupswas calculated from 3D images using deformation based morphometry. Neurodevelopmental assessmentswere obtained from preterm infants at a mean chronological age of 27.5 months, and from controls at a meanage of 31.1 months.We identified a common image phenotype in 66 of 80 preterm infants at term equivalent comprising: diffusewhite matter injury; and tissue volume reduction in the dorsomedial nucleus of the thalamus, the globuspallidus, periventricular white matter, the corona radiata and within the central region of the centrumsemiovale (t=4.42 pb0.001 false discovery rate corrected). The abnormal image phenotype is associatedwith reduced median developmental quotient (DQ) at 2 years (DQ=92) compared with control infants(DQ=112), pb0.001.These findings indicate that specific neural systems are susceptible to maldevelopment after preterm birth,and suggest that neonatal image phenotype may serve as a useful biomarker for studying mechanisms ofinjury and the effect of putative therapeutic interventions.

Simpson Centre for Reproduc-nce Crescent, Edinburgh EH16

rdman).

l rights reserved.

t al., A common neonatal image phenotype), doi:10.1016/j.neuroimage.2010.04.261

© 2010 Elsevier Inc. All rights reserved.

Introduction

The prevalence of preterm birth is increasing in resource-richcountries (Goldenberg et al., 2008), and this presents a growingburden to education services, as well as affected individuals and theirfamilies, because neurocognitive impairment occurs in 40–50% ofchildren born extremely preterm (Bhutta et al., 2002; Larroque et al.,2008; Marlow et al., 2005). The neural substrates that underlieimpairment are not fully understood, but diffuse white matter injuryon magnetic resonance (MR) imaging is seen commonly amongpreterm infants at term equivalent age, and is likely to contribute tothe substrate for impairment because it occurs with a similar

prevalence, and is associated with short term measures of adverseneurodevelopmental outcome (Dyet et al., 2006; Krishnan et al.,2007). Previous imaging studies suggest that whitematter injury doesnot occur in isolation, but rather, it is associated with maldevelop-ment of remote grey matter structures, which suggests thatdeveloping neural systems are affected by preterm birth (Inderet al., 1999; Lin et al., 2001).

We previously carried out a group-wise comparison of volumechange maps from preterm infants and term-born controls withrespect to a common template, and detected discreet areas of volumereduction in the thalami and lentiform nuclei in the preterm group,whichwasmost apparent among those infants with diffuse non-cysticwhite matter abnormalities (Boardman et al., 2006). However, thechildren were too young for functional correlates of the neonatalimage phenotype to be assessed, and the random field family-wiseerror rate (FWER) correction for multiple tests was used, which isnow known to be a less powerful approach comparedwith techniques

predicts adverse neurodevelopmental outcome in

2 J.P. Boardman et al. / NeuroImage xxx (2010) xxx–xxx

ARTICLE IN PRESS

that control the false discovery rate (the proportion of false rejectionsof the null hypothesis among the total number of rejections), whenrelatively small groups are studied (Benjamini and Hochberg, 1995;Genovese et al., 2002; Laird et al., 2005; Langers et al., 2007). In thisstudy we test the hypothesis that after premature birth there is animaging phenotype comprised of white matter abnormality and focalmorphological change which predicts neurodevelopmental impair-ment in early childhood.

Materials and methods

Participants

The MR images of 80 preterm infants at term equivalent age(33 males and 47 females), and 20 term-born control infants (12 malesand 8 females) were analyzed. The infants were recruited at Hammer-smith Hospital between February 2001 and November 2003. Preterminfants were eligible for recruitment if theywere born at≤34 completedweeks of gestation and had no congenital malformation or congenitalcentral nervous system infection. Infants with cystic periventricularleucomalacia, hemorrhagic parenchymal infarction or post-hemorrhagicventricular dilation were excluded. All control infants were consideredhealthy and recruited from the post-natal wards.

The preterm group underwent MR imaging at a mean of 40+3 weekspost-menstrual age (range 37+6–43+6), and the control group under-wentMR imaging at amean age of 40+3 weeks PMA (range 36+4–43+1).Demographic and anthropometric details are shown in Table 1. Twenty-six of the preterm infants were growth restricted [birth weight b10thcentile for age, Child Growth Foundation software (Freeman et al., 1995)],and five required supplemental oxygen at 36 weeks post-menstrual age.No infant received post-natal steroids.

Preterm infants were sedated for the MR examination with chloralhydrate, and control infants were examined in natural sleep. Pulseoximetry, electrocardiographic and televisual monitoring were usedthroughout the examination. Ear protection was used (Natus Mini-Muffs, Natus Medical Inc., San Carlos, CA). Ethical approval wasprovided by the Hammersmith and Queen Charlotte's and ChelseaResearch Ethics Committee.

Image acquisition

A Philips 1.5 T MR system was used to acquire high resolution T1weighted (TR=30ms, TE=4.5 ms, flip angle=30°) volume datasets incontiguous slices with a voxel size of 1.0×1.0×1.6 mm, as well astransverse T1-weighted conventional spin echo (TR 500/TE 15 ms) andT2-weighted fast-spin echo (TR 4500/TEeff 210 ms). DWI was acquiredusing a single shot EPI sequence with the following parameters: TR6000 ms, TE 100ms, 100×100 matrix, FOV 24 cm, slice thickness 5 mm.A reference image was obtained with a b value of 0 (nominal value), andDWIswere obtainedwith a b value of 1000 s/mm2 in the read, phase andslice directions.

Table 1Neonatal demographic and anthropometric characteristics with preterm group segregated

Preterm infants at ter

Group A

n=66

PMA at birth/weeks (mean and range) 29+6 (24+1–33+5)Birth weight/g (mean and range) (SDS [SD]) 1259 (610–2226) (−Occipitofrontal head circumference at birth/cm(mean and range) (SDS [SD])

27.0 (22.0–32.5) (−

Weight at time of image acquisition/g (mean and range)(SDS [SD])

3144 (1700–4380) (−

Occipitofrontal head circumference at scan/cm (mean and range)(SDS [SD])

35.2 (31.2–38.0) (0.0

Please cite this article as: Boardman, J.P., et al., A common neonatal imchildren born preterm, NeuroImage (2010), doi:10.1016/j.neuroimage.2

Apparent diffusion coefficient (ADC) values were calculated bypositioning ROIs in frontal, central and posterior white matter at thelevel of the centrum semiovale on the reference image (b=0) and on theread, phase and slice DWIs, using published methods (Counsell et al.,2003). Infantswere classified ashavingwhitematter injury if oneormorewhitematter region had an ADC value greater than 2 standard deviationsabove the mean of that measured in a group of normal term controls:N1.566×10−3 mm2/s (frontal white matter); N1.378×10−3 mm2/s(central white matter); N1.474×10−3 mm2/s (posterior white matter)(Boardman et al., 2006).

Deformation-based morphometry

A high dimensional non-rigid registration algorithm which uses afree form deformation model (cubic B-spline) and normalised mutualinformation as the maximised similarity measure was used to bring allsubjects into alignment with a reference template, chosen as the 3Dimage of an infant born at term (Rueckert et al., 1999; Rueckert et al.,2003; Boardman et al., 2006). The goal of the registration process is toachieve precise spatial correspondence between all subjects and theanatomy of the reference subject. The output of the registration processis a deformation field from which voxel-wise spatially resolved scalarmeasurements of volume change are extracted (Davatzikos et al., 1996;Studholme et al., 2004). A qualitative evaluation of the accuracy ofanatomical alignment of the transformed imageswith the templatewasmade before deformation fields were used to calculate volume changes.

Statistical analysis

Volume change maps for each subject relative to the referenceanatomy were analyzed using a two-sample t-test implemented inSPM5 (http://www.fil.ion.ucl.ac.uk/spm/) to identify regions ofstatistically significant difference in tissue volume between groups.Analyses were carried out using control of the false discovery ratewith a significance threshold of pb0.001. Only voxel clusters N10were considered in the model. The effect of quantitatively defineddiffuse white matter injury on tissue volume was assessed bycomparing preterm infants with and without white matter injury toterm-born controls.

Neurodevelopmental outcome

Neurodevelopmental progress was assessed using the GriffithsMental Development Scales (Revised) (Huntley, 1996) which providean overall developmental quotient (DQ) with subscales assessing skillareas (locomotor, personal–social, hearing and language, eye and handcoordination, performance, and additionally practical reasoning forchildren N2 years of age). The developmental and subscale quotientsfrom the Griffiths scales are calculated taking into account the exact ageof the child at the time of the assessment. Scores are presented forchronological age because the majority of infants were over 2 years at

by image phenotype.

m equivalent age Term-born controls n=20

Group B

n=14

30 (24+5–34) 39+2 (35+4–42)0.70 [1.06]) 1307 (640–2190) (−0.81 [1.21]) 3381 (2448–4780) (−0.16 [1.00])0.64 [1.05]) 27.4 (24.8–30.6) (−0.67 [1.67]) 34.9 (33.0–38.0) (−0.1 [1.27])

0.94 [1.36]) 3095 (2170–3560) (−1.34 [0.86]) 3381 (2448–4780) (0.13 [1.06])

8 [1.49]) 34.8 (33.0–38.0) (−0.43 [1.21]) 34.9 (33.0–38.0) (−0.1 [1.27])

age phenotype predicts adverse neurodevelopmental outcome in010.04.261

3J.P. Boardman et al. / NeuroImage xxx (2010) xxx–xxx

ARTICLE IN PRESS

the time of assessment, and because the Griffiths scores calculated onchronological age are more predictive of functional outcome in laterchildhood than scores corrected for prematurity (van den Hout et al.,1998; Miller et al., 1984). All children were examined neurologicallylooking specifically for evidence of cerebral palsy (Bax et al., 2005).All assessments were carried out by experienced developmentalpaediatricians unaware of the quantitative MRI findings.

Weight and head circumference were measured and the standarddeviation scores (SDS) with respect to the age and sex adjustedmeanswere calculatedwith the Child Growth Foundation software (Freemanet al., 1995).

The relationship between developmental quotient and gestationalage at birth was examined by estimating the linear regressioncontrolling for age at time of assessment. Parametric tests were usedfor all analyses, except the analysis of variance of the DQ scores at2 years because these values were not normally distributed, andrequired a non-parametric rank based method. All analyses werecarried out using SPSS v16.0.

Results

Computational anatomic phenotype associated with diffuse whitematter injury

Of the 80 preterm infants that had DWI images free of movementartefact and suitable for quantitative analysis, 66 had abnormally highADC values in at least one white matter region at the level of thecentrum semiovale (group A). Fourteen had ADC values withinnormal limits throughout the white matter (group B). There was nosignificant difference between preterm groups A and Bwith respect togestational age at birth (p=0.66) or birth weight (p=0.68).

Fig. 1. Focal tissue volume reduction associated with diffuse white matter injury. Focal vocomplex of the thalami, the globus pallidi and the posterior periventricular white matter (Fpreterm infants at term equivalent age who have diffuse white matter injury compared withno significant clusters of voxels throughout the volume among preterm infants with normsignificance threshold (Figs. 1D–F).

Please cite this article as: Boardman, J.P., et al., A common neonatal imchildren born preterm, NeuroImage (2010), doi:10.1016/j.neuroimage.2

Group A had discrete tissue loss within the dorsomedial complexof the thalami, the globus pallidi, and focally within the posteriorperiventricular white matter, the corona radiata and the centrumsemiovale compared with term controls compared with controls(t=4.42 pb0.001, FDR). No morphological changes were detected ingroup B at the same significance threshold (Fig. 1) compared tocontrols. There were no significant volumetric differences betweenpreterm groups A and B when multiple tests were controlled.

Neurodevelopmental outcome

Sixty-nine percent (55/80) of the preterm group were assessed at amean chronological age of 27.5 months (range 24.0–31.0). The meanweight of the preterm group at follow-up was 12.94 kg (range 9.18–17.64) with mean SDS (weight) 0.29 (SD 1.12); the mean OFC of thepreterm group was 49.1 cm (44.8–54.0), with mean SDS (OFC)−0.45(SD 1.24). Therewas no significant difference in gestational age at birth,prevalence of supplemental oxygendependency or IUGRbetween thosewhowere seenat2 years and thosewhowerenot. Eighty-fivepercentofthe term control group (17/20) were assessed at a mean age of31.1 months (range 23.5–40.0). None of the children had cerebral palsy.

The DQ values were not normally distributed: the median DQ for allpreterm infants was 101 (range 68–141). The median DQ of the term-born control group was 112 (range 94–129). There was a dose-dependent relationship between DQ and gestational age at birth(B=1.9, 95% CI 0.353–2.026, p=0.006) (Fig. 2). Age at time ofassessment was not significant in the model.

The DQ and subscale quotients for preterm group A, preterm groupB, and for the controls are shown in Table 2: there was significantinequality in themedian DQ between preterm group A, preterm groupB and term controls, Table 3 (H=18.825, pb0.001 Kruskal–Wallis).

lume reduction in the posterior part of the corona radiata (Fig. 1A), the dorsomedialig. 1B), and focally within the central region of the centrum semiovale (Fig. 1C) amongterm controls (t=4.42, pb0.001 FDR corrected, shown in highlighted red). There wereal ADC values in white matter (group B) compared with term controls at the same

age phenotype predicts adverse neurodevelopmental outcome in010.04.261

Fig. 2. Neurodevelopmental outcome and prematurity at birth. There is a significantrelationship between gestational age at birth and developmental quotient (B=1.38,95% CI 0.464–2.303, p=0.004, linear regression). The fit line with 95% confidenceinterval for the population is displayed.

Table 3Kruskal–Wallis one way analysis of variance for median DQ at 2 years based on pretermimage phenotype and control patients (H=18.825, d.f. 2, pb0.001).

Median DQ (range) Mean rank

Preterm group A 92 (59–131) 29.18Preterm group B 102 (64–112) 35.69Term controls 112 (94–129) 54.59

4 J.P. Boardman et al. / NeuroImage xxx (2010) xxx–xxx

ARTICLE IN PRESS

Post-hoc tests show that this is accounted for by a highly significantdifference in DQ between preterm group A (pb0.01) and controls,whereas the difference in DQ scores between preterm group B andcontrols (p=0.054), and between group A and group B (p=0.354)did not reach statistical significance (Mann–Whitney test withBonferroni correction for multiple tests).

The significant effect attributable to neonatal image phenotyperemainedwhen DQ scores were corrected for degree of prematurity atbirth H=8.85, p=0.012.

Discussion

We identified a common highly significant anatomic phenotypeconsisting of diffuse white matter injury and focal tissue loss localised tothe dorsomedial nucleus of the thalamus, the globus pallidus, and withinthe white matter of the corona radiata, posterior periventricular whitematter, and the central region of the centrum semiovale. The abnormalneonatal image phenotype was associated with a lower DQ at 2 years ofage compared to preterm infants without this pattern of injury and toterm controls. Our data support the hypothesis that white matter injuryimpacts upon the development of basal ganglia and thalami, and that thisdisturbance to the connectivity of developing neural systems hasimportant functional consequences. This pattern of injury in survivinginfants is consistent with neuropathological studies of infants who died,which show involvement of both grey and white matter structures(recently labelled the ‘encephalopathy of prematurity’), and give confi-dence that these post-mortem data truly reflect the effects of prematurityrather than the events associated with mortality (Volpe, 2009). We have

Table 2Griffiths Mental Development Scales at around 2 years chronological age for preterm infan

DQ(SD)

Locomotor(SD)

Personal–social(SD)

Hearing and la(SD)

Preterm group A(n=46/66)

92 (16) 98 (16) 97 (13) 93 (18)

Preterm group B(n=8/14)

102 (17) 97 (13) 102 (16) 95 (16)

Term controls(n=17/20)

112 (11) 115 (13) 116 (13) 117 (18)

a Practical reasoning was only assessed in children older than 24 months at the time of a

Please cite this article as: Boardman, J.P., et al., A common neonatal imchildren born preterm, NeuroImage (2010), doi:10.1016/j.neuroimage.2

shown previously that these structural changes are not attributable to aglobal failure in brain growth (Boardman et al., 2007).

Focal tissue volume changes were not detected in preterm infantswithout white matter injury (group B) with respect to the controls,but neither was difference detected between the two preterm groups.This may be attributable to type 2 error because preterm group Bconsisted of only 14 subjects. Further investigation of a larger group ofpreterm infants with normal diffusion parameters in white matterwould be required to address this issue. Unfortunately, priorspecification of the required sample size is not possible in the absenceof established methods for power calculations in this type ofmorphometric study. However it is predictable that very large groupswould be needed because the general effect of prematurity (as seen inTables 2 and 3, and Fig. 2) on neural function means that even thoseinfants with less severe white matter injury not detected as anabnormal ADC might have some lesser element of deep grey matterchange. A design that included region of interest (ROI) measurementsand prior hypotheses about effect size could enable accurate samplesize calculations, but robust tools to segment the ROIs identified inthis study are not yet available.

We present neurodevelopmental outcome scores calculated onchronological age at time of assessment because most of the childrenwere older than 24months, and although there is no universal agreementthere is some consensus to cease correction at this age (Johnson andMarlow, 2006). Specifically, with respect to the Griffiths scales,uncorrected DQ scores are probably more predictive of cognitiveperformance at 5.5 years (van den Hout et al., 1998). One limitation ofthe study is that outcome data were only available for 69% of the pretermgroup. It is possible thatwesawchildrenwhowerenot representative, butthere was no significant difference in known neonatal risk factors foradverse outcome between the two preterm groups.

The dorsomedial nucleus of the thalamus consists of threemorphologically distinct regions (magnocellular, parvocellular, anddensocellular portions), and is the principal thalamic relay nucleus forthe prefrontal cortex, as well as receiving afferent connections from thestriatum, basal ganglia and limbic system. It has reciprocal projections tothe prefrontal cortex and also projects to anterior cingulate, anteriorinsular and dorsomedial frontal cortices (Goldman-Rakic and Porrino,1985; Jones, 1997; Armstrong, 1990). This connectivity distributionimplies a role in integrating cognition with affective experience, andneuronal number in the nucleus appears to be associated withbehavioural function and disease in adults (Jelsing et al., 2006; Younget al., 2004).

ts segregated by image phenotype and for term controls.

nguage Eye and hand coordination(SD)

Performance(SD)

aPractical reasoning(SD)

89 (10) 97 (21) 110 (21)a

96 (9) 107 (19) 112 (12)b

99 (14) 110 (17) 116 (11)c

ssessment (a=26, b=5, c=14).

age phenotype predicts adverse neurodevelopmental outcome in010.04.261

5J.P. Boardman et al. / NeuroImage xxx (2010) xxx–xxx

ARTICLE IN PRESS

The nucleus has a unique developmentally regulated pattern ofneuronal death and survival in that there is a relative excess inneuronal number in the human newborn compared with the adult(Abitz et al., 2007). This could be explained by a second wave ofneuronalmigration from the telencephalic ganglionic eminence to thedorsal thalamus after the earlier migration of neurones from theventricular zone (Letinic and Rakic, 2001; Bystron et al., 2008). Theloss of volume in this region that we detected in association withpremature birth and diffuse white matter injury could represent anattenuation of the second wave of neurones from the telencephalon,deafferentation secondary to white matter injury or cortical abnor-malities, or dysregulation of the apoptotic mechanisms that governthalamic development (Volpe, 2009). It is relevant that in a post-mortem series of infants who died with cystic PVL the dorsomedialnucleus was especially affected by neuronal loss over and above otherthalamic nuclei (Ligam et al., 2008).

The globus pallidus consists of an internal (GPi) and external (GPe)segment separated by themedial medullary lamina. Themain afferentinput is from the dorsal striatum and substantia nigra, and the GPisends inhibitory connections to the thalamus. The globus pallidus hasbeen identified as an essential substrate for controlling access toworking memory (McNab and Klingberg, 2008), which underpins avariety of cognitive functions including general intelligence, attentionand reasoning (Suss et al., 2002; Conway et al., 2003), and which isimpaired among children born preterm (Woodward et al., 2005). Ourdata are consistent with a disturbance to this system contributing tolater functional impairment in preterm infants.

We identified tissue contraction within white matter amongchildren with abnormally high ADC values. Volume reduction wasdistributed within the posterior periventricular white matter, and isconsistent with the relative enlargement of the adjacent posteriorhorns of the lateral ventricles seen in the preterm population(Peterson et al., 2003; Boardman et al., 2006). Other foci of reducedwhite matter volume were detected in the corona radiata and themiddle region of the centrum semiovale (pb0.001). Since these areamong the first white matter regions to myelinate (Counsell et al.,2002), it is possible that focal reductions in tissue volume may reflectreduced or delayed myelination in these tracts compared with infantsborn at term. Maldevelopment of posterior periventricular whitematter would be expected to result in motor impairment, andalthough none of the children in this study had cerebral palsy,changes in this region could explain the reduction in locomotor scoresseen in the preterm group compared with controls (Table 2), andcontribute to the substrate to minor motor impairment that iscommon among children born preterm (Marlow et al., 2007).

Cortical volume reduction at term equivalent age has beenreported when global tissue classification methods are used (Inderet al., 2005).We did not expect to detect global cortical changes in thisapplication of DBM because it is sensitive to local change that isspatially consistent across the whole group, rather than summatedchanges detected bymethods that use tissue classification.We did notdetect cerebellar volume loss; this might be because cerebellargrowth appears to be preserved in the absence of major destructivesupratentorial lesions (Limperopoulos et al., 2005; Srinivasan et al.,2006) and we excluded such infants from our study.

A potential limitation of the study was classification of diffuse whitematter injury based on ADC values. We chose this metric because itprovides an objective measure of diffuse white matter injury (Counsellet al., 2003), but measures of white matter microstructure that takeaccount of anisotropy, such as fractional anisotropy (FA), may provide amore refinedmetric ofwhitematter integrity. Regional or tract based FAis likely to be helpful for investigating the growth of specific structureswith respect to their connectivity patterns.

Identifying a statistically significant neonatal MR phenotype withfunctional correlates suggests that early image acquisition withcomputational analysis is valuable for providing objective data on

Please cite this article as: Boardman, J.P., et al., A common neonatal imchildren born preterm, NeuroImage (2010), doi:10.1016/j.neuroimage.2

neurodevelopmental prognosis for preterm infants. Morphometryadvances understanding of the disturbances to neural systems thatoccur in association with premature birth and it may be a usefulbiomarker for investigating mechanisms of injury and the efficacy ofinterventions designed to reduce the burden of cognitive impairmentfollowing preterm birth.

Acknowledgments

This work was supported by the Imperial College HealthcareBiomedical Research Centre funding scheme; the Medical ResearchCouncil; Philips Medical Systems; and the Engineering and PhysicalSciences Research Council and the Garfield Weston Foundation. Weare grateful to the families who chose to take part in the study, and tostaff of Queen Charlotte's and Chelsea and Hammersmith Hospitalswho helped to facilitate imaging and neurodevelopmentalassessments.

References

Abitz, M., Nielsen, R.D., Jones, E.G., Laursen, H., Graem, N., Pakkenberg, B., 2007. Excessof neurons in the human newborn mediodorsal thalamus compared with that ofthe adult. Cereb. Cortex 17, 2573–2578.

Armstrong, E., 1990. The limbic thalamus: anterior andmediodorsal nuclei. In: Paxinos, G.(Ed.), TheHumanNervous System. Academic Press, Inc., San Diego (CA), pp. 469–478.

Bax, M., Goldstein, M., Rosenbaum, P., Leviton, A., Paneth, N., Dan, B., Jacobsson, B.,Damiano, D., 2005. Proposed definition and classification of cerebral palsy, April2005. Dev. Med. Child Neurol. 47, 571–576.

Benjamini, Y., Hochberg, Y., 1995. Controlling the false discovery rate: a practical andpowerful approach to multiple testing. J. R. Stat. Soc. B Methodol. 57, 289–300.

Bhutta, A.T., Cleves, M.A., Casey, P.H., Cradock, M.M., Anand, K.J., 2002. Cognitive andbehavioral outcomes of school-aged children who were born preterm: a meta-analysis. JAMA 288, 728–737.

Boardman, J.P., Counsell, S.J., Rueckert, D., Hajnal, J.V., Bhatia, K.K., Srinivasan, L.,Kapellou, O., Aljabar, P., Dyet, L.E., Rutherford, M.A., Allsop, J.M., Edwards, A.D.,2007. Early growth in brain volume is preserved in the majority of preterm infants.Ann. Neurol. 62, 185–192.

Boardman, J.P., Counsell, S.J., Rueckert, D., Kapellou, O., Bhatia, K.K., Aljabar, P., Hajnal, J.,Allsop, J.M., Rutherford, M.A., Edwards, A.D., 2006. Abnormal deep grey matterdevelopment following preterm birth detected using deformation-based mor-phometry. Neuroimage 32, 70–78.

Bystron, I., Blakemore, C., Rakic, P., 2008. Development of the human cerebral cortex:Boulder Committee revisited. Nat. Rev. Neurosci. 9, 110–122.

Conway, A.R., Kane, M.J., Engle, R.W., 2003. Working memory capacity and its relationto general intelligence. Trends Cogn. Sci. 7, 547–552.

Counsell, S.J., Allsop, J.M., Harrison, M.C., Larkman, D.J., Kennea, N.L., Kapellou, O.,Cowan, F.M., Hajnal, J.V., Edwards, A.D., Rutherford, M.A., 2003. Diffusion-weightedimaging of the brain in preterm infants with focal and diffuse white matterabnormality. Pediatrics 112, 1–7.

Counsell, S.J., Maalouf, E.F., Fletcher, A.M., Duggan, P., Battin, M., Lewis, H.J., Herlihy, A.H., Edwards, A.D., Bydder, G.M., Rutherford, M.A., 2002. MR imaging assessment ofmyelination in the very preterm brain. AJNR Am. J. Neuroradiol. 23, 872–881.

Davatzikos, C., Vaillant, M., Resnick, S.M., Prince, J.L., Letovsky, S., Bryan, R.N., 1996.A computerized approach for morphological analysis of the corpus callosum. J.Comput. Assist. Tomogr. 20, 88–97.

Dyet, L.E., Kennea, N., Counsell, S.J., Maalouf, E.F., Ajayi-Obe, M., Duggan, P.J., Harrison,M., Allsop, J.M., Hajnal, J., Herlihy, A.H., Edwards, B., Laroche, S., Cowan, F.M.,Rutherford, M.A., Edwards, A.D., 2006. Natural history of brain lesions in extremelypreterm infants studied with serial magnetic resonance imaging from birth andneurodevelopmental assessment. Pediatrics 118, 536–548.

Freeman, J.V., Cole, T.J., Chinn, S., Jones, P.R.,White, E.M., Preece,M.A., 1995. Cross sectionalstature and weight reference curves for the UK, 1990. Arch. Dis. Child. 73, 17–24.

Genovese, C.R., Lazar, N.A., Nichols, T., 2002. Thresholding of statistical maps infunctional neuroimaging using the false discovery rate. Neuroimage 15, 870–878.

Goldenberg, R.L., Culhane, J.F., Iams, J.D., Romero, R., 2008. Epidemiology and causes ofpreterm birth. Lancet 371, 75–84.

Goldman-Rakic, P.S., Porrino, L.J., 1985. The primate mediodorsal (MD) nucleus and itsprojection to the frontal lobe. J. Comp. Neurol. 242, 535–560.

Huntley, M., 1996. The Griffiths Mental Development Scales: from birth to 2 years(revised). Association for Research in Infant and Child Development (ARICD).

Inder, T.E.,Huppi, P.S.,Warfield, S., Kikinis, R., Zientara,G.P., Barnes, P.D., Jolesz, F., Volpe, J.J.,1999. Periventricular white matter injury in the premature infant is followed byreduced cerebral cortical gray matter volume at term. Ann. Neurol. 46, 755–760.

Inder, T.E., Warfield, S.K., Wang, H., Huppi, P.S., Volpe, J.J., 2005. Abnormal cerebralstructure is present at term in premature infants. Pediatrics 115, 286–294.

Jelsing, J., Hay-Schmidt, A., Dyrby, T., Hemmingsen, R., Uylings, H.B., Pakkenberg, B.,2006. The prefrontal cortex in the Gottingen minipig brain defined by neuralprojection criteria and cytoarchitecture. Brain Res. Bull. 70, 322–336.

Johnson, S., Marlow, N., 2006. Developmental screen or developmental testing? EarlyHum. Dev. 82, 173–183.

age phenotype predicts adverse neurodevelopmental outcome in010.04.261

6 J.P. Boardman et al. / NeuroImage xxx (2010) xxx–xxx

ARTICLE IN PRESS

Jones, E.G., 1997. Thalamic organization and chemical anatomy. In: Jones, M., Jones, E.G.,McCormick, D.A. (Eds.), Thalamus, Vol. I: Organisation and Function. ElsevierScience Ltd., Oxford, pp. 31–126.

Krishnan, M.L., Dyet, L.E., Boardman, J.P., Kapellou, O., Allsop, J.M., Cowan, F., Edwards,A.D., Rutherford, M.A., Counsell, S.J., 2007. Relationship between white matterapparent diffusion coefficients in preterm infants at term-equivalent age anddevelopmental outcome at 2 years. Pediatrics 120, e604–e609.

Laird, A.R., Fox, P.M., Price, C.J., Glahn, D.C., Uecker, A.M., Lancaster, J.L., Turkeltaub, P.E.,Kochunov, P., Fox, P.T., 2005. ALEmeta-analysis: controlling the false discovery rateand performing statistical contrasts. Hum. Brain Mapp. 25, 155–164.

Langers, D.R., Jansen, J.F., Backes, W.H., 2007. Enhanced signal detection in neuroimagingbymeans of regional control of the global false discovery rate. Neuroimage 38, 43–56.

Larroque, B., Ancel, P.Y., Marret, S., Marchand, L., Andre,M., Arnaud, C., Pierrat, V., Roze, J.C., Messer, J., Thiriez, G., Burguet, A., Picaud, J.C., Breart, G., Kaminski, M., 2008.Neurodevelopmental disabilities and special care of 5-year-old children born before33 weeks of gestation (the EPIPAGE study): a longitudinal cohort study. Lancet 371,813–820.

Letinic, K., Rakic, P., 2001. Telencephalic origin of human thalamic GABAergic neurons.Nat. Neurosci. 4, 931–936.

Ligam, P., Haynes, R.L., Folkerth, R.D., Liu, L., Yang, M., Volpe, J.J., Kinney, H.C., 2008.Thalamic damage in periventricular leukomalacia: novel pathologic observationsrelevant to cognitive deficits in survivors of prematurity. Pediatr. Res. 65, 524–529.

Limperopoulos, C., Soul, J.S., Haidar, H., Huppi, P.S., Bassan, H., Warfield, S.K., Robertson,R.L., Moore, M., Akins, P., Volpe, J.J., du Plessis, A.J., 2005. Impaired trophicinteractions between the cerebellum and the cerebrum among preterm infants.Pediatrics 116, 844–850.

Lin, Y., Okumura, A., Hayakawa, F., Kato, K., Kuno, T., Watanabe, K., 2001. Quantitativeevaluation of thalami and basal ganglia in infants with periventricular leukoma-lacia. Dev. Med. Child Neurol. 43, 481–485.

Marlow, N., Hennessy, E.M., Bracewell, M.A., Wolke, D., 2007. Motor and executivefunction at 6 years of age after extremely preterm birth. Pediatrics 120, 793–804.

Marlow, N., Wolke, D., Bracewell, M.A., Samara, M., 2005. Neurologic and developmentaldisability at six years of age after extremely preterm birth. N. Engl. J. Med. 352, 9–19.

McNab, F., Klingberg, T., 2008. Prefrontal cortex and basal ganglia control access toworking memory. Nat. Neurosci. 11, 103–107.

Please cite this article as: Boardman, J.P., et al., A common neonatal imchildren born preterm, NeuroImage (2010), doi:10.1016/j.neuroimage.2

Miller, G., Dubowitz, L.M., Palmer, P., 1984. Follow-up of pre-term infants: is correctionof the developmental quotient for prematurity helpful? Early Hum. Dev. 9,137–144.

Peterson, B.S., Anderson, A.W., Ehrenkranz, R., Staib, L.H., Tageldin, M., Colson, E., Gore, J.C., Duncan, C.C., Makuch, R., Ment, L.R., 2003. Regional brain volumes and their laterneurodevelopmental correlates in term and preterm infants. Pediatrics 111,939–948.

Rueckert, D., Frangi, A.F., Schnabel, J.A., 2003. Automatic construction of 3-D statisticaldeformation models of the brain using nonrigid registration. IEEE Trans. Med.Imaging 22, 1014–1025.

Rueckert, D., Sonoda, L.I., Hayes, C., Hill, D.L., Leach, M.O., Hawkes, D.J., 1999. Nonrigidregistration using free-form deformations: application to breast MR images. IEEETrans. Med. Imaging 18, 712–721.

Srinivasan, L., Allsop, J., Counsell, S.J., Boardman, J.P., Edwards, A.D., Rutherford, M.,2006. Smaller cerebellar volumes in very preterm infants at term-equivalent ageare associated with the presence of supratentorial lesions. AJNR Am. J. Neuroradiol.27, 573–579.

Studholme, C., Cardenas, V., Blumenfeld, R., Schuff, N., Rosen, H.J., Miller, B., Weiner, M.,2004. Deformation tensor morphometry of semantic dementia with quantitativevalidation. Neuroimage 21, 1387–1398.

Suss, H.-M., Oberauer, K., Wittmann, W.W., Wilhelm, O., Schulze, R., 2002. Working-memory capacity explains reasoning ability—and a little bit more. Intelligence 30,261–288.

van den Hout, B.M., Eken, P., Van der, L.D., Wittebol-Post, D., Aleman, S., Jennekens-Schinkel, A., van der Schouw, Y.T., de Vries, L.S., van, N.O., 1998. Visual, cognitive,and neurodevelopmental outcome at 51/2 years in children with perinatalhaemorrhagic–ischaemic brain lesions. Dev. Med. Child Neurol. 40, 820–828.

Volpe, J.J., 2009. Brain injury in premature infants: a complex amalgam of destructiveand developmental disturbances. Lancet Neurol. 8, 110–124.

Woodward, L.J., Edgin, J.O., Thompson, D., Inder, T.E., 2005. Object working memorydeficits predicted by early brain injury and development in the preterm infant.Brain 128, 2578–2587.

Young, K.A., Holcomb, L.A., Yazdani, U., Hicks, P.B., German, D.C., 2004. Elevated neuronnumber in the limbic thalamus in major depression. Am. J. Psychiatry 161,1270–1277.

age phenotype predicts adverse neurodevelopmental outcome in010.04.261