SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting Information

Aim

Preterm birth is associated with an increased risk of adverse neurodevelopmental outcomes. Diffusion magnetic resonance imaging (dMRI) combined with tractography can be used to assess non-invasively white matter microstructure and brain development in preterm infants. Our aim was to conduct a systematic review of the current evidence obtained from tractography studies of preterm infants in whom MRI was performed up to term-equivalent age.

Method

Databases were searched for dMRI tractography studies of preterm infants.

Results

Twenty-two studies were assessed. The most frequently assessed tracts included the corticospinal tract, the corpus callosum, and the optic radiations. The superior longitudinal fasciculus, and the anterior and superior thalamic radiations were investigated less frequently. A clear relationship exists between diffusion metrics and postmenstrual age at the time of scanning, although the evidence of an effect of gestational age at birth and white matter injury is conflicting. Sex and laterality may play an important role in the relationship between diffusion metrics, early clinical assessment, and outcomes.

Interpretation

Studies involving infants of all gestational ages are required to elucidate the relationship between gestational age and diffusion metrics, and to establish the utility of tractography as a predictive tool. There is a need for more robust acquisition and analysis methods to improve the accuracy of assessing development of white matter pathways.

Abbreviations
ASD

Autism spectrum disorder

BSID-III

Bayley Scales of Infant Development III

CSF

Cerebrospinal fluid

CST

Corticospinal tract

dMRI

Diffusion magnetic resonance imaging

DTI

Diffusion tensor imaging

PLIC

Posterior limb of the internal capsule

PMA

Post-menstrual age

SLF

Superior longitudinal fasciculus

WMI

White matter injury

Preterm births account for approximately 12% of live births in the USA[1] and are associated with an increased risk of infant mortality, morbidity, and neurodevelopmental disabilities. Neurodevelopmental disabilities include motor impairments such as cerebral palsy (CP) and mild motor dysfunction, as well as cognitive, language, and behavioural disorders and a higher rate of attention-deficit–hyperactivity disorder (ADHD) and autism spectrum disorder (ASD).[2] Advanced brain imaging techniques, such as magnetic resonance imaging (MRI), enable the non-invasive assessment of brain maturation and development in preterm infants.[3]

Diffusion MRI offers unique insights into the microstructure and organization of white matter by probing the random motion of water molecules. This random motion is hindered or restricted in the direction perpendicular to white matter tracts, even in unmyelinated tracts.[4, 5] Microstructural properties of white matter are often described using quantitative diffusion tensor imaging (DTI)-derived metrics: fractional anisotropy and mean diffusivity.[4] Fractional anisotropy is a quantitative, unitless measure that describes the degree of anisotropy of diffusion. Fractional anisotropy ranges from 0 to 1, with a value of 0 indicating perfectly isotropic diffusion and a value approaching 1 indicating increased anisotropic diffusion. Low values of fractional anisotropy are typically observed within the cerebrospinal fluid (CSF), whereas high values of fractional anisotropy are found in highly organized white matter. Mean diffusivity, on the other hand, describes the overall diffusion in units of 10–3mm2/s. In contrast to fractional anisotropy, the values of mean diffusivity are high in regions of unrestricted diffusion (such as CSF) and lower in regions of restricted diffusion (such as white matter). Increased fractional anisotropy and decreased mean diffusivity are, therefore, typically associated with higher organization, (pre-) myelination and decreased water content.[6] Changes in organization, (pre-) myelination and water content during early brain development make these metrics ideal candidates for the assessment of maturation.

Diffusion MRI also offers the possibility of non-invasive delineation of white matter pathways using tractography by following the direction of preferred water diffusion. The delineated pathways can subsequently be used as three-dimensional regions of interest and metrics, such as fractional anisotropy or mean diffusivity within the tracts, can be explored.

The aim of this systematic review was to evaluate information obtained from dMRI tractography studies of preterm infants, with a specific focus on the effect of postmenstrual age (PMA) at the time of MRI, gestational age at birth, sex, laterality, and white matter injury (WMI) on fractional anisotropy and mean diffusivity within tracts. This review will provide information on important confounding factors that need to be adjusted in statistical analysis, and identify areas for future research.

Method

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting Information

Search terms and inclusion criteria

A literature search of PubMed, Web of Knowledge, Scopus, and EMBASE was conducted on 11 March 2013 for the keywords ‘preterm’ or ‘premature’ (and variations thereof), and ‘tractography’, ‘fiber tracking’, or ‘fibre tracking’. Only peer-reviewed publications in English were considered.

To be included, a study had to meet the following criteria: (1) it was an original research article (i.e. excluding review articles and conference abstracts), (2) at least one study group consisted of preterm infants who underwent MRI at or before term-equivalent age, and (3) diffusion tractography was performed. Studies using tractography only to confirm the location of otherwise obtained regions, with no assessment of the delineated tracts themselves, were excluded. Studies performing tractography post mortem or in utero, as well as case studies, were also excluded. Two independent raters (KP and SMS) assessed eligibility based on the title and abstract of the article. After the identification of eligible articles, references therein were also screened for eligibility.

Data extraction and analyses

The following information was extracted from all included studies: (1) the number of participants; (2) the gestational age range if provided, otherwise the mean and standard deviation or median and interquartile range; (3) the PMA range if provided, otherwise the mean and standard deviation or the median and interquartile range; and (4) the white matter tracts that were investigated. Studies were subsequently grouped by investigated tract(s) as well as the study results regarding the effect of (1) PMA, (2) gestational age, (3) sex, (4) hemisphere, and (5) WMI on diffusion metrics and (6) the relationship between diffusion metrics and neurodevelopmental outcome and categorized as ‘association found’, ‘no association found’ or ‘not assessed’. Only diffusion metrics fractional anisotropy and mean diffusivity were considered in this review; other metrics, including axial and radial diffusivity, tract length, and volume, were assessed in a smaller number of included studies and are not reported here.

Results

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting Information

Search results

The results of the retrieval process that was used to identify studies are reported in Figure 1. The literature databases PubMed, Web of Knowledge, Scopus, and EMBASE yielded 126 results after removal of duplicates, which were assessed independently by two reviewers (KP and SMS). Of the 64 original research articles in the English language identified from the search, 43 were excluded on the basis of the title and abstract (reasons for exclusion are given in Fig. 1). The remaining 21 papers[7-27] met our full inclusion criteria. One additional publication[28] was identified from the references of eligible studies. Included studies are presented in Table 1.

Table 1. Details of included studies. Studies are grouped by identical acquisition protocols (see Table SI for acquisition details)
StudySample sizeGestational age at birth (wks)Postmenstrual age at imaging (wks)TractsMetricsCrossing fibre model?Tractography: deterministic or probabilistic?
  1. Note that all figures are given in the format as shown in the original publication. a–jStudies with the same superscript indicator have identical acquisition protocols. kMean (standard deviation). lSixty-nine infants scanned before 46 weeks PMA. mMedian (interquartile range). CST, corticospinal tract; FA, fractional anisotropy; MD, mean diffusivity; λ1, primary eigenvalue; λ2, intermediate eigenvalue; λ3, minimum eigenvalue; STR, superior thalamic radiation; λ23, average of λ2 and λ3; OR, optic radiation; CC, corpus callosum; FOF, fronto-occipital fasciculus; SLF, superior longitudinal fasciculus; ATR, anterior thalamic radiation; PTR, posterior thalamic radiation; TC, thalamocortical; Cl, case linear anisotropy.

Partridge[8]a9 preterm, 5 serially25–3433–39, 35–43CSTFA, MD, λ1, λ2, λ3NoDeterministic
Berman[8]a27 preterm, 10 seriallyNG28–43CST, STRFA, MD, λ1, λ23NoDeterministic, Probabilistic
Berman[9]a36 preterm20.3–33.129–41ORFA, MD, λ1, λ23NoDeterministic
Glass[12]a9 preterm, 6 serially26+3–32+329+1–34+5, 32+6–40+1ORFA, MD, λ1, λ23NoDeterministic
Thompson[19]b106 preterm, 22 term27.6 (1.7)k, 38.8 (1.1)k38–42CCFA, MD, λ1, λ23, volumeNoProbabilistic
Thompson[26]b106 preterm27.6 (1.7)k38–42CCFA, MD, λ1, λ23, volumeNoProbabilistic
Yoo[7]c6 preterm26–3028–40CC, CST, OR, FOFNoneNoDeterministic
Adams[13]d55 preterm, all serially24–3227.4–40.4, 34.3–46.4CSTFA, MD, λ1, λ23NoDeterministic
Zwicker[14]d126 preterm, all serially24–3232.1 (30.5–33.9)m, 40.3 (38.7–42.6)mCSTFA, MD, λ1, λ23NoDeterministic
Bassi[15]e37 preterm24+1–32+340+1–43ORFANoProbabilistic
Bassi[21]e46 preterm25+1–35+239–43+6CC, CST, OR, SLFFANoProbabilistic
Hasegawa[20]f58 preterm23.00–33.7137.43–43.71CCFA, MDNoDeterministic
Aeby[10]g22 preterm, 6 term26.7–32, 37.8–40.634–41CC, CST, ATR, STR, PTRFAYesProbabilistic
Liu[16]g27 preterm26.0–34.435.4–42.1CST, ATR, STR, PTR, SLFFA, MD, λ1, λ23, volumeYesProbabilistic
Liu[17]g38 preterm26.0–34.435.4–42.0CST, ATR, STR, PTR, SLFFA, MD, λ1, λ23, volumeYesProbabilistic
Liu[24]g70 preterm26.0–36.035.4–42.9CC, CST, ATR, STR, PTR, SLFFA, MD, λ1, λ23, volumeYesProbabilistic
Ball[25]h47 preterm, 18 term23+4–34+6, 36+0–41+638+2–44+4TCAnisotropy/ CoherenceYesProbabilistic
Groppo[28]h53 preterm, 22 serially25+4–34+629+5–36, 39+2–46CST, ORFAUnknownProbabilistic
van Kooij[22]i67 preterm28.6 (1.8)k41.5 (1.1)kCC, CSTFA, MD, Cl, λ1, λ23, volume, lengthNoDeterministic
van Pul[23]i89 preterm28.5 (1.7)k39.6–44.7CC, CSTFA, MD, Cl, volume, lengthNoDeterministic
de Bruïne[18]j84 preterm25.6–31.940.0–62.1lCC, CSTFA, MD, lengthNoDeterministic
de Bruïne[27]j64 preterm25.6–31.940–62CC, CSTFA, MD, lengthNoDeterministic
image

Figure 1. Process of study selection.

Download figure to PowerPoint

Study characteristics

Information regarding number of participants, gestational age at birth, PMA at time of imaging, assessed tracts, and metrics is summarized in Table 1. Information regarding MRI acquisition parameters are provided in Table SI. Participant numbers range from 6 to 126 (median 49), and gestational age from as early as 24+1 weeks. MRI data were acquired in the first weeks of life (i.e. at any age from birth to term equivalent) in four studies;[7-10] serially in five studies;[11-14, 28] and around term age in 13 studies.[15-27] The corticospinal tract (CST) was studied most frequently,[7, 10, 11, 13, 14, 16-18, 21-24, 27, 28] followed by the corpus callosum[7, 10, 18-24, 26, 27] and the optic radiations/posterior thalamic radiations.[7, 9, 10, 12, 15-17, 21, 24, 28]

The majority of studies used the diffusion tensor model for tractography[7-9, 11-15, 18-23, 26, 27] only a small number of studies employed higher-order models of diffusion to enable the resolution of crossing fibres for improved tractography accuracy.[10, 16, 17, 24, 25] Probabilistic and deterministic tractography techniques were used in approximately equal numbers of studies (Table 1).

Tracts of interest

Tractography of the preterm infant brain has focused on a small number of major white matter tracts, including the CST, the corpus callosum, the superior longitudinal fasciculus, and the thalamocortical connections, including the anterior, superior, and posterior thalamic (optic) radiations (Fig. 2). Table 2 details the studies that have assessed the relationship between fractional anisotropy and/or mean diffusivity of tracts and PMA, gestational age, time of life, differences between preterm and term infants, sex, laterality, and WMI.

Table 2. Studies assessing associations between diffusion measures and clinical parameters
 Association found?Postmenstrual ageGestational age Extra-uterine lifePreterm–termLateralitySexWhite matter injuryEarly assessment (vision)Prediction
  1. aTotal number of results may be higher than total number of papers when more than one tract was investigated in a paper. CST, corticospinal tract; CC, corpus callosum; RB, rostral body; AMB, anterior midbody; PMB, posterior midbody; IST, isthmus; SPL, splenium; SLF, superior longitudinal fasciculus; ATR, anterior thalamic radiation; HOM, higher-order model of diffusion (anisotropy/coherence); STR, superior thalamic radiation; PTR, posterior thalamic radiation.

Fractional anisotropy
CST (n=12)Yes7 (Partridge,[11] Berman,[8] Aeby,[10] Adams,[13] de Bruïne,[18] van Pul,[23] Groppo[28])2 (van Pul,[23] Groppo[28])1 (Groppo[28])2 (Liu,[16] van Pul[23])1 (Adams[13])2 (van Kooij,[22] de Brüine[27])
No3 (Adams,[13] Liu,[16] de Bruïne[18])1 (de Bruïne[18])3 (Liu,[17] van Kooij,[22] Zwicker[14])1 (van Pul[23])
CC whole (n=6)Yes2 (Aeby,[10] van Pul[23])2 (van Pul,[23] Thompson[26])1 (Thompson[26])
No1 (van Pul[23])1 (Thompson[19])1 (van Pul[23])1 (Liu[24])1 (van Kooij[22])
CC genu (n=5)Yes1 (Thompson[26])
No1 (de Bruïne[18])2 (de Bruïne,[18] Hasegawa[20])1 (Thompson[19])2 (de Bruïne,[18] Bassi[21])1 (Thompson[26])
CC RB (n=2)Yes1 (Thompson[26])
No1 (Thompson[19])1 (Thompson[26])
CC AMB (n=4)Yes1 (Thompson[26])
No1 (Hasegawa[20])1 (Thompson[19])2 (Thompson,[26] de Bruïne[27])
CC PMB (n=2)Yes1 (Thompson[26])
No1 (Thompson[19])1 (Thompson[26])
CC IST (n=4)Yes1 (Thompson[26])
No1 (Hasegawa[20])1 (Thompson[19])1 (de Bruïne[18])1 (Thompson[26])
CC SPL (n=6)Yes1 (de Bruïne[18])1 (Hasegawa[20])1 (Thompson[26])
No1 (Thompson[19])1 (de Bruïne[18])2 (Thompson,[26]) de Bruïne[27])
SLF (n=4)Yes1 (Liu[16])
No1 (Liu[17])2 (Bassi,[21] Liu[24])
ATR (n=5)Yes1 (Aeby[10])1 (Ball[25]) (HOM)
No– (Liu[16])– (Liu[17])– (Liu[24])
STR (n=6)Yes2 (Berman,[8] Aeby[10])1 (Ball[25]) (HOM)1 (Liu[24])
No1 (Liu[16])1 (Liu[17])
PTR (n=10)Yes3 (Berman,[9] Aeby,[10] Groppo[28])1 (Groppo[28])1 (Groppo[28])1 (Ball[25]) (HOM)3 (Bassi,[15] Berman,[9] Groppo[28])1 (Glass[12])
No1 (Liu[16])1 (Liu[17])2 (Bassi,[21] Liu[24])1 (Groppo[28])
Mean diffusivity
CST (n=11)Yes5 (Partridge,[11] Berman,[8] Adams,[13] de Bruïne,[18] van Pul[23])1 (van Pul[23])1 (Liu[16])2 (Adams,[13] van Pul[23])1 (van Kooij[22])
No3 (Adams,[13] Liu,[16] de Bruïne[18])2 (van Pul,[23] de Bruïne[18])3 (Liu,[17] van Kooij,[22] Zwicker[14])2 (de Bruïne,[18] Liu[24])1 (de Bruïne[27])
CC whole (n=5)Yes1 (van Pul[23])1 (Thompson[19])1 (van Pul[23])1 (Thompson[26])
No1 (van Pul[23])1 (van Pul[23])1 (Liu[24])1 (van Kooij[22])
CC genu (n=4)Yes1 (de Bruïne[18])1 (Thompson[19])1 (Thompson[26])
No1 (de Bruïne[18])1 (de Bruïne[18])2 (Thompson,[26] de Bruïne[27])
CC RB (n=2)Yes1 (Thompson[19])
No1 (Thompson[26])1 (Thompson[26])
CC AMB (n=2)Yes1 (Thompson[19])1 (Thompson[26])
No1 (Thompson[26])
CC PMB (n=2)Yes1 (Thompson[19])1 (Thompson[26])
No1 (Thompson[26])
CC IST (n=2)Yes1 (Thompson[19])1 (Thompson[26])
No1 (Thompson[26])
CC SPL (n=4)Yes1 (de Bruïne[18])1 (Thompson[19])1 (Thompson[26])2 (Thompson,[26] de Bruïne[27])
No1 (de Bruïne[18])
SLF (n=3)Yes
No1 (Liu[16])1 (Liu[17])1 (Liu[24])
ATR (n=3)Yes1 (Liu[24])
No1 (Liu[16])1 (Liu[17])
STR (n=3)Yes1 (Berman[8])1 (Liu[24])
No1 (Liu[16])1 (Liu[17])
PTR (n=5)Yes1 (Berman[9])1 (Glass[12])
No1 (Liu[16])1 (Liu[17])1 (Liu[24])1 (Berman[9])
Summary
Fractional anisotropy
Total papers 8 61 234 83 5 
Total resultsa (association reported/not found)17 (16/1)12 (4/8)2 (2/0)10 (3/7)7 (3/4)8 (0/8)25 (13/12)3 (3/0)14 (4/10) 
Mean diffusivity
Total papers 6 40 134 51 4 
Total resultsa (association reported/not found)10 (10/0) 6 (1/5)0 (0/0) 7 (0/7)7 (1/6)8 (0/8)18 (10/8)1 (0/1)13 (5/8) 
image

Figure 2. Tracts assessed by included studies. Shown are the corpus callosum (CC) in its subdivisions and as a whole; the corticospinal tracts; the superior longitudinal fasciculus; and the anterior, superior, and posterior thalamic radiations.

Download figure to PowerPoint

Corticospinal tract

The CST (Fig. 2) is the major motor projection that descends from the motor cortex, through the posterior limb of the internal capsule (PLIC), to the brainstem. It is a large pathway that is readily delineated using tractography. Preterm birth is a risk factor for the development of CP and milder motor impairments and, for these reasons, the CST was the most frequently assessed white matter tract in preterm infants, both during the first weeks of life[7, 8, 10] and at term-equivalent age,[16-18, 21-24, 27] as well as serially at both time points.[11, 13, 14, 28]

Age at scan

The effect of PMA on fractional anisotropy and mean diffusivity has been assessed for infants scanned during the first weeks of life,[8, 10] for infants scanned around term-equivalent age,[18, 23] and for infants scanned serially.[11, 13, 28] These studies consistently show an increase in fractional anisotropy[8, 10, 11, 18, 23, 28] and a decrease in mean diffusivity[8, 11, 13, 18, 23] with increasing PMA. Additionally, Berman et al.[8] assessed fractional anisotropy values within each axial slice of the CST, and reported that a ‘dip’ in fractional anisotropy within the centrum semiovale can be observed in participants of higher PMA that was not present when they were younger. The appearance of this ‘dip’ is likely to be the result of the increasing presence of crossing fibres in this area. Groppo et al.[28] reported an additional effect of the time of extrauterine life on fractional anisotropy within the CST.

Degree of preterm birth

The effect of gestational age at birth on diffusion metrics in the CST is less clear. Three studies[13, 16, 18] report no relationship between gestational age at birth and fractional anisotropy or mean diffusivity when MRI was performed serially[13] or around term age.[16, 18] One study performed around term age[23] found a decrease in fractional anisotropy and an increase in mean diffusivity in the left CST, but not in the right CST, with increasing gestational age at birth. In contrast, another study in which infants were scanned serially[28] reported an increase in fractional anisotropy with increasing gestational age at birth.

Laterality

In neonatal studies, values obtained from the left and right CST are frequently averaged. A few studies of infants scanned only around term age have assessed the validity of this approach,[16, 18, 23] and these have shown inconclusive results. Two studies[16, 23] reported higher fractional anisotropy in the left CST, while a third study[18] found no difference in fractional anisotropy of the left and right CST. One study[16] showed lower mean diffusivity in the left CST, while the other two studies[18, 23] found no evidence of mean diffusivity asymmetry.

Sex

The effect of sex on DTI metrics of the CST has not been extensively studied in preterm infants, but has mostly been studied at term age[17, 22, 23] and serially in one study.[14] No study reported any differences between males and females in fractional anisotropy or mean diffusivity of the CST, or in the rate of change in these metrics with PMA. There is evidence to suggest that the relationship between these parameters and neurodevelopmental outcome may depend on sex[22] (see below).

White matter injury

The effect of WMI on fractional anisotropy or mean diffusivity of the CST was assessed at term age in four studies,[18, 21, 23, 24] and serially in one study.[13] Adams et al.[13] reported a slower rate of increase in fractional anisotropy with increasing PMA in infants with moderate to severe WMI than in infants with no or only mild WMI, although the rate of decrease in mean diffusivity with increasing PMA did not differ between groups. They also reported a higher mean diffusivity for any given PMA in infants with moderate to severe WMI, and lower fractional anisotropy in infants with severe WMI.[13] The study by Bassi et al.[21] showed reduced fractional anisotropy in the CST of infants with punctate lesions and a relationship between fractional anisotropy and lesion load. A relationship between increasing WMI score and decreasing mean diffusivity, but no relationship between WMI score and fractional anisotropy, was shown by van Pul et al.[23] (n=89). The two remaining studies did not find any differences in fractional anisotropy or mean diffusivity between infants with no, mild, or moderate WMI[24] (scored using the scheme of Woodward et al.[29]), and in fractional anisotropy between infants with normal–mild, moderate, and severe WMI[18] (scored using the scheme proposed by Leijser et al.[30]) respectively.

Neurodevelopmental outcome

Two studies assessed the relationship between diffusion metrics of the CST obtained around term age and neurodevelopmental outcome at 2 years of age.[22, 27] In the first study,[22] outcome was assessed with the cognitive, fine, and gross motor aspects of the Bayley Scales of Infant Development III, 3rd edition (BSID-III[31]). The authors found no relationship between outcome and fractional anisotropy or mean diffusivity when the group was studied as a whole. Different associations with outcome did, however, emerge when males and females were assessed in separate groups: in males, better fine motor outcome was related to increased fractional anisotropy within the left CST.[22] In the second study,[27] cognitive and psychomotor development was assessed using BSID-III, and children with CP were identified using the Gross Motor Function Classification System (GMFCS). It was found that lower fractional anisotropy was a predictor of delayed psychomotor development and CP.[27]

Corpus callosum

The corpus callosum (Fig. 2) is the largest white matter bundle that connects the left and right hemispheres of the brain. The corpus callosum plays an essential role in the interhemispheric transfer of motor, sensory, and higher cognitive information. Preterm birth is frequently associated with thinning of the corpus callosum. Tractography has been used to study the corpus callosum as a whole or in segments in preterm infants during the first weeks of life[7, 10] and at term-equivalent age,[18-24, 26, 27] but not serially.

Age at scan

Few studies have investigated the effect of PMA on corpus callosum development during the first weeks of life[10] or around term age.[18, 23] Two studies[10, 23] assessed the corpus callosum as a whole and found that fractional anisotropy increases with increasing PMA,[10, 23] while mean diffusivity decreases.[23] A third study[18] investigated the effect of PMA on the development of corpus callosum tracts passing through the genu and splenium. Fractional anisotropy was related to PMA in the splenium, but not in the genu, and mean diffusivity was related to PMA in both the genu and the splenium.[18]

Degree of preterm birth

The effect of preterm birth on corpus callosum development has also not been extensively assessed. van Pul et al.[23] assessed the corpus callosum as a whole around term age and found no association between gestational age and fractional anisotropy or mean diffusivity. Similarly, de Bruïne et al.[18] found no association between fractional anisotropy or mean diffusivity and gestational age in the genu and the splenium of the corpus callosum. Hasegawa et al.[20] on the other hand, found that fractional anisotropy was significantly lower in the splenium tracts in infants born at <25 weeks, or between 26 and 29 weeks, than in infants born between 30 and 32 weeks, while there were no group differences in the body and genu tracts. They also reported a relationship between gestational age and fractional anisotropy in splenium, but not in isthmus, tracts. Finally, Thompson et al.[26] found no differences in fractional anisotropy between preterm and term infants in the whole corpus callosum or any of its subdivisions (genu, rostral body, anterior midbody, posterior midbody, isthmus, splenium), while mean diffusivity was increased in the whole corpus callosum and all of its subdivisions.

Sex

Only a single study assessed sex effects on diffusion of the corpus callosum as a whole,[23] and reported no differences in fractional anisotropy or mean diffusivity between the sexes.

White matter injury

Three studies investigated the effect of WMI on the corpus callosum when assessed as a whole[23, 24, 26] or in its subdivisions.[18, 21, 26] When assessing the corpus callosum as a whole, two studies[23, 26] reported decreased fractional anisotropy and increasing mean diffusivity with an increasing WMI severity score.[29, 32] In contrast, Liu et al.[24] found no difference in fractional anisotropy or mean diffusivity between infants with mild WMI and infants without WMI. Thompson et al.[26] also reported a decrease in fractional anisotropy with increasing WMI severity in all subdivisions of the corpus callosum, and an increase in mean diffusivity in all subdivisions except the rostral body. Contrary to these findings, de Bruïne et al.[18] found no evidence of alterations in fractional anisotropy or mean diffusivity within the genu and splenium of the corpus callosum between infants with normal/mild, moderate, and severe WMI. Finally, Bassi et al.[21] identified no change in fractional anisotropy in infants with punctate lesions compared with infants without punctate lesions.

Outcome

Neurodevelopmental outcome at 2 years of age has been assessed in relation to the corpus callosum in three studies.[22, 26, 27] van Kooij et al.[22] reported no relationship between 2-year outcome (measured using the cognitive, gross, and fine motor aspects of BSID-III[31]) and fractional anisotropy or mean diffusivity of the corpus callosum as a whole around term age when the entire group was assessed, or when males and females were assessed separately. Thompson et al.[26] on the other hand, report that a higher mental development index at 2 years of age (measured using the mental developmental index of BSID-II[33]) is related to higher fractional anisotropy values within the whole corpus callosum and the anterior midbody. Furthermore, a higher psychomotor developmental index (measured using BSID-II[33]) was related to lower mean diffusivity in the splenium.[26] This finding is supported by the results of de Bruïne et al.[27] who reported that reduced mean diffusivity in the splenium is a predictor for delayed psychomotor development.

Superior longitudinal fasciculus

The superior longitudinal fasciculus (SLF; Fig. 2) is a bilateral association pathway that connects the frontal, occipital, and temporal lobes. It consists of four distinct components in the human brain that are bundled together.[34] It is believed that functions of the SLF include regulation of motor behaviour, perception of visual space, transmission of auditory spatial information, and language articulation.[34] In preterm infants, the SLF has not been studied in great detail, and only at term-equivalent age.[16, 17, 21]

Age at scan

The effect of PMA on diffusion and tractography characteristics has not been studied.

Degree of preterm birth

The effect of gestational age on diffusion and tractography characteristics has not been studied.

Laterality

Liu et al.[16] reported higher fractional anisotropy in the left parietotemporal SLF, with no lateralization of mean diffusivity in this tract. No effect of laterality was found for the frontoparietal SLF.[16]

Sex

Liu et al.[17] identified a trend towards lower mean diffusivity in the left parietotemporal SLF in females, with no difference in fractional anisotropy. No effect of sex was found for the frontoparietal SLF.[17]

White matter injury

Diffusion or tractography metrics of the SLF are not altered by the presence of punctate lesions[21] or mild white matter abnormalities (WMA).[24]

Thalamic connections

The thalamus is a deep grey matter structure which relays motor and sensory information to the cortex. The anterior, superior, and posterior thalamic radiations (which include the optic radiations) have been assessed in preterm infants. One study used an original strategy to investigate the thalamocortical connectome in an automated fashion.[25]

Optic radiations (posterior thalamic radiations)

The posterior thalamic radiations (Fig. 2), which include the optic radiations, have been assessed in preterm infants during the first weeks of life,[7, 9, 10] at term age,[15-17, 21, 24, 25] and serially.[12, 28]

Age at scan

In preterm infants scanned at 30 weeks PMA or younger, the optic radiations could not be delineated in an early study.[7] Several studies have reported increasing fractional anisotropy with increasing PMA.[9, 10, 28] A single study assessed mean diffusivity and found a negative relationship with PMA.[9]

Degree of preterm birth

Groppo et al.[28] reported an increase in fractional anisotropy with increasing gestational age, as well as a relationship between the time of extrauterine life and fractional anisotropy.[28] Ball et al.[25] showed reduced connectivity (a measure that includes both anisotropy and fibre coherence[35]) of the optic radiations in preterm infants at term age compared with term infants.

Laterality

The only study assessing asymmetry of the posterior thalamic radiations[16] reported no asymmetry in fractional anisotropy or mean diffusivity between hemispheres.

Sex

Liu et al.[17] reported no effect of sex on fractional anisotropy or mean diffusivity of the optic radiations. No other study has assessed the effect of sex.

Retinopathy of prematurity

Bassi et al.[15] found no evidence that retinopathy of prematurity affects fractional anisotropy within the optic radiations at term age.

White matter injury

Two studies assessed the effect of WMI on diffusion of the posterior thalamic radiations.[21, 24] No differences in fractional anisotropy between infants with or without punctate lesions were identified.[21] Fractional anisotropy and mean diffusivity were not affected by mild WMAs.[24]

Early visual assessments

Three studies[9, 15, 28] consistently reported a decrease in fractional anisotropy with impaired visual function when visual assessment was performed within days of MRI assessment. Early MRI, however, did not predict visual function at term age.[28]

Visual outcome

Using visual evoked potentials, Glass et al.[12] showed that fractional anisotropy and mean diffusivity were related to the peak response amplitudes for spatial frequency sweeps, but not to contrast or vernier offset.

Superior thalamic radiations

The superior thalamic radiations (Fig. 2), containing motor and sensory components, were assessed before[8, 10] and at term-equivalent age.[16, 17, 24, 25]

Age at scan

An increase in fractional anisotropy with increasing PMA was noted.[8, 10] In addition, decreased mean diffusivity was found.[8] Berman et al.[8] also reported the appearance of a ‘dip’ in fractional anisotropy in the centrum semiovale in older infants that was not present in younger infants.

Degree of preterm birth

Ball et al.[25] found a significant reduction in connectivity (a measure that includes both anisotropy and fibre coherence[35]) of the superior thalamic radiations in preterm infants compared with term-born infants.

Laterality

No differences in fractional anisotropy or mean diffusivity between hemispheres were observed for the motor or sensory superior thalamic radiation in a single study.[16]

Sex

No sex effect on fractional anisotropy and mean diffusivity of the motor and sensory superior thalamic radiation were reported in a single study.[17]

White matter injury

Liu et al.[24] found lower fractional anisotropy in the left sensory superior thalamic radiation, as well as higher mean diffusivity in the left sensory and bilateral motor superior thalamic radiation in infants with mild WMA compared with infants without WMA.

Anterior thalamic radiation

The anterior thalamic radiations (Fig. 2) were investigated during the first weeks of life,[10] and at term age.[16, 17, 24, 25]

Age at scan

Aeby et al.[10] reported an increase in fractional anisotropy with increasing PMA.

Degree of term birth

Decreased connectivity (a measure that includes both anisotropy and fibre coherence[35]) in preterm infants compared with term-born infants has been reported.[25]

Laterality

Liu et al.[16] found no asymmetry between hemispheres in fractional anisotropy or mean diffusivity.

Sex

Liu et al.[17] found no effect of sex on fractional anisotropy or mean diffusivity.

White matter injury

While no difference[24] between infants with mild WMA (n=27) and infants without WMA(n=41) was found for fractional anisotropy, mean diffusivity was increased in the left anterior thalamic radiation in infants with mild WMA.[24]

Discussion

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting Information

In this systematic review, we have summarized the main findings of tractography studies performed in preterm infants. Our analysis revealed that assessment is restricted to a small number of major white matter pathways, namely the CST, the corpus callosum, and optic radiations, as well as, less frequently, the SLF and anterior and superior thalamic radiations. In this review, we have collated the evidence of the impact of PMA at the time of MRI, gestational age at birth, postnatal age, sex, laterality, and WMI on fractional anisotropy and mean diffusivity.

Age at MRI and degree of prematurity

The effect of PMA at the time of MRI on fractional anisotropy and mean diffusivity measures has been well established. With the exception of a single study,[18] all studies have reported an increase in fractional anisotropy and a decrease in mean diffusivity with increasing PMA. These changes are thought to reflect ongoing processes of organization, (pre-)myelination, and changes in brain water content.[36] The effect of the degree of preterm birth (i.e. gestational age at birth) on diffusion metrics is, however, less clear. The majority of studies reported no relationship between gestational age and fractional anisotropy or mean diffusivity for the majority of assessed tracts. Only three studies[20, 23, 28] identified relationships between gestational age and fractional anisotropy or mean diffusivity, but the relationships were contradictory: a positive relationship between fractional anisotropy and gestational age and a negative relationship between mean diffusivity and gestational age were found in two studies,[20, 28] whereas a negative relationship between fractional anisotropy and gestational age, and a positive relationship between mean diffusivity and gestational age was reported in one study.[23] Two studies have compared diffusion metrics obtained from white matter pathways of preterm infants scanned around term age with those of term-born infants.[19, 25] Their findings suggest that, although fractional anisotropy of the corpus callosum is unaffected by the degree of preterm birth,[19] both the mean diffusivity of the corpus callosum[19] and a higher-order diffusion anisotropy measure of thalamocortical connections[25] may be affected by preterm birth. A potential additional confounding factor of diffusion metrics is the postnatal age.[28]

It is clear that the relationship between gestational age and microstructure needs to be established with more certainty, which will require both a comparison of preterm infants at term age and healthy term-born infants as well as scanning a larger number of infants of all gestational ages. Research has thus far focused primarily on very preterm and extremely preterm infants, thereby reducing the power to establish a relationship between gestational age and microstructure given the relatively narrow gestational age range studied.

Sex

The three studies that have assessed the effect of sex on diffusion measures of the CST,[17, 22] the corpus callosum,[23] the SLF,[17] and thalamic radiations[17] found no effect on fractional anisotropy or mean diffusivity. It has, however, been shown that different relationships exist between diffusion measures and neurodevelopmental outcome in males compared with females.[22] In males there seems to be a disadvantage with respect to mortality, morbidity, and neurodevelopmental outcomes including an increased risk of CP. The reasons for this sex dimorphism are unclear.[37]

Given the small number of studies that have assessed sex effects, which are important in some key aspects of brain development, further research is warranted to clarify any effect of sex on microstructure with tractography. Data from males and females should be analysed separately or sex should be considered as a potential confounding factor in statistical analysis.

Lateralization

The effect of laterality on diffusion metrics in preterm infants is not yet well understood. Although Liu et al.[16] reported increased fractional anisotropy and decreased mean diffusivity in the left CST compared with the right CST, van Pul et al.[23] reported increased fractional anisotropy in the left CST with increased GA, but not in the right CST; moreover, de Bruïne et al.[18] found no effect of hemisphere on either fractional anisotropy or mean diffusivity of the CST. Liu et al.[16] also identified an effect of hemisphere on fractional anisotropy, but not mean diffusivity, of the SLF, and no effect of hemisphere on fractional anisotropy or mean diffusivity of thalamic radiations. Further research is warranted to establish the potential importance of lateralization of motor and language pathways in preterm infants and the subsequent clinical correlates.

White matter injury

The effect of WMI on diffusion metrics remains unclear, with some studies finding an association[13, 21, 23, 24, 26] and others not[18, 21, 23, 24, 26] for all assessed pathways. A possible reason for these discrepancies may be differences in the definition[29, 30, 32, 38] and severity of WMI in the various studies, as well as differences in statistical analysis. One study assessed group differences between infants with and without punctate lesions,[21] while other studies assessed group differences between infants with mild WMI and no WMI,[24] infants with low or high WMI scores,[26] infants with normal-to-mild, moderate, and severe WMI,[18] or correlations between diffusion metrics and WMI scores.[13, 23]

The development of robust standardized quantitative measures of WMI and their adoption would increase interpretability, comparability, and repeatability of future studies.

Early neurodevelopment and neurodevelopmental outcome

The relationship between diffusion metrics and concurrent early developmental scores has been assessed only for the optic radiations/posterior thalamic radiations,[9, 15, 28] with evidence suggesting that fractional anisotropy of the optic radiations is related to visual function at the time of MRI. The predictive value of diffusion metrics using tractography is not yet fully established, and whether a relationship with long-term outcome is found varies by tract, outcome measure, time of MRI, and time of outcome assessment.[12, 22, 26-28] For example, both van Kooij et al.[22] and Thompson et al.[26] examined 2-year outcome in relation to fractional anisotropy and mean diffusivity within the entire corpus callosum around term age in large cohorts (n=67 and n=106, respectively), but reported on different outcome measures. van Kooij et al.[22] found no relationship with cognitive, gross, and fine motor aspects of BSID-II,[33] and Thompson et al.[26] reported a relationship between fractional anisotropy and cognitive development, and between mean diffusivity and psychomotor development (both measured using BSID-III[31]).

In order to clearly define the relationship between diffusion measures of the various tracts and specific aspects of functional neurodevelopmental outcome, further investigation is required. The use of whole brain network-based techniques,[39] which do not require strong a priori hypotheses regarding tracts of interest, may provide new insights. The addition of further early neurological assessments with good test characteristics for prognostication of outcome, such as general movements,[40] as well as the use of serial MRI to obtain a more ‘dynamic’ picture of maturation in the individual infant, may lead to improved prediction of neurodevelopmental outcomes.

Limitations of tractography studies

Compared with studies performed in children or adults, diffusion MRI of preterm and term-born infants is challenging owing to smaller head size, increased (and time-varying) brain water content, and increased risk of movement artefact. In a recent review,[41] we highlighted the difficulties in acquiring, preprocessing, and analysing diffusion MRI of the neonatal brain, and described strategies to help improve data quality. These include acquiring more diffusion encoding directions than is required for the reconstruction model in order to introduce redundancy and enable the rejection of individual image volumes that may be affected by head movement. They also include additional correction for image distortions, inherent in the imaging technique (e.g. distortions caused by eddy currents and susceptibility inhomogeneities), and image artefacts caused by head movement and cardiac pulsation.

The majority of studies assessed in this review have relied on the diffusion tensor model to delineate white matter pathways (Table 1). The major limitation of the diffusion tensor model is its inability to resolve crossing fibres. A number of higher-order models of diffusion have been developed to overcome this limitation (see Tournier et al.[42] for a review). The use of these higher-order models of diffusion significantly reduces the incidence of false positives (i.e. delineation of pathways known not to exist) and false negatives[43] (i.e. failure to delineate existing pathways). Few tractography studies of preterm infants have used these improved techniques.[10, 16, 17, 24, 25] Furthermore, interpretation of the quantitative metrics obtained using the diffusion tensor model (such as fractional anisotropy, as well as axial and radial diffusivity) in crossing fibre regions is difficult, and counterintuitive results have been reported.[44] Diffusion metrics based on a higher-order model of diffusion have also recently been developed,[45, 46] but only one study has used a diffusion metric based on a higher-order model of diffusion with tractography in preterm infants.[25] These novel metrics, as well as assessment of metrics along the length of the tract,[47] may provide novel insights into the development of the preterm brain.

It is essential that future diffusion MRI and tractography studies of the preterm infant brain should be based on advanced models of diffusion[42] and follow a rigorous pre-processing and quality assurance pipeline. Serial imaging of infants will facilitate the understanding of maturational changes in the individual infant and enhance prognostic capability. The usefulness of diffusion MRI may be enhanced if it is accompanied by other quantitative MRI techniques, such as myelin water fraction and relaxation time mapping,[48] perfusion imaging,[49] and detailed early neurological assessments.[50, 51]

Conclusion

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting Information

Only a relatively small number of white matter pathways have been investigated in preterm infants, with studies focusing mostly on the CST, the corpus callosum, and optic radiations. The SLF, and the anterior and superior thalamic radiations have been investigated less frequently. A clear relationship exists between diffusion metrics and PMA, while the evidence of an effect of gestational age and WMI is conflicting. Sex and laterality may play an important role in the relationship between diffusion metrics and early clinical assessment, as well as outcomes. Studies employing a higher-order model of diffusion, and involving a large number of infants of all gestational ages, are required to elucidate the relationship between gestational age and diffusion metrics, and to establish the utility of tractography as a tool for the prediction of neurodevelopmental outcome. Improved prediction of outcome may be achieved by combining information obtained from high-quality diffusion tractography and early neurological assessments.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting Information

The authors received funding support from The University of Queensland (International Postgraduate Research Scholarship [KP], Centennial Scholarship [KP], Research Scholarship [SMS]) and the Australian National Health and Medical Research Council (Career Development Fellowship [RNB: 10037220], Practitioner Fellowship [PBC: 511117]).

References

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting Information
  • 1
    Martin JA, Hamilton BE, Ventura SJ, Osterman MJK, Wilson EC, Mathews TJ. Births: final data for 2010. Natl Vital Stat Rep 2012; 61: 172.
  • 2
    Behrman RE, Butler AS. Preterm Birth: Causes, Consequences, and Prevention. Washington, DC: National Academies Press, 2007.
  • 3
    Mathur AM, Neil JJ, Inder TE. Understanding brain injury and neurodevelopmental disabilities in the preterm infant: the evolving role of advanced magnetic resonance imaging. Semin Perinatol 2010; 34: 5766.
  • 4
    Mukherjee P, McKinstry RC. Diffusion tensor imaging and tractography of human brain development. Neuroimaging Clin N Am 2006; 16: 1943.
  • 5
    Dudink J, Kerr JL, Paterson K, Counsell SJ. Connecting the developing preterm brain. Early Hum Dev 2008; 84: 77782.
  • 6
    Hüppi PS, Dubois J. Diffusion tensor imaging of brain development. Semin Fetal Neonatal Med 2006; 11: 48997.
  • 7
    Yoo SS, Park HJ, Soul JS, et al. In vivo visualization of white matter fibre tracts of preterm- and term-infant brains with diffusion tensor magnetic resonance imaging. Invest Radiol 2005; 40: 1105.
  • 8
    Berman JI, Mukherjee P, Partridge SC, et al. Quantitative diffusion tensor MRI fibre tractography of sensorimotor white matter development in premature infants. Neuroimage 2005; 27: 86271.
  • 9
    Berman JI, Glass HC, Miller SP, et al. Quantitative fibre tracking analysis of the optic radiation correlated with visual performance in premature newborns. AJNR Am J Neuroradiol 2009; 30: 1204.
  • 10
    Aeby A, Liu Y, De Tiège X, et al. Maturation of thalamic radiations between 34 and 41 weeks’ gestation: a combined voxel-based study and probabilistic tractography with diffusion tensor imaging. AJNR Am J Neuroradiol 2009; 30: 17806.
  • 11
    Partridge SC, Mukherjee P, Berman JI, et al. Tractography-based quantitation of diffusion tensor imaging parameters in white matter tracts of preterm newborns. J Magn Reson Imaging 2005; 22: 46774.
  • 12
    Glass HC, Berman JI, Norcia AM, et al. Quantitative fibre tracking of the optic radiation is correlated with visual-evoked potential amplitude in preterm infants. AJNR Am J Neuroradiol 2010; 31: 14249.
  • 13
    Adams E, Chau V, Poskitt KJ, Grunau RE, Synnes A, Miller SP. Tractography-based quantitation of corticospinal tract development in premature newborns. J Pediatr 2010; 156: 8828.
  • 14
    Zwicker JG, Grunau RE, Adams E, et al. Score for Neonatal acute physiology-II and neonatal pain predict corticospinal tract development in premature newborns. Pediatr Neurol 2013; 48: 1239.
  • 15
    Bassi L, Ricci D, Volzone A, et al. Probabilistic diffusion tractography of the optic radiations and visual function in preterm infants at term-equivalent age. Brain 2008; 131: 57382.
  • 16
    Liu Y, Balériaux D, Kavec M, et al. Structural asymmetries in motor and language networks in a population of healthy preterm neonates at term-equivalent age: a diffusion tensor imaging and probabilistic tractography study. Neuroimage 2010; 51: 7838.
  • 17
    Liu Y, Metens T, Absil J, et al. Sex differences in language and motor-related fibres in a population of healthy preterm neonates at term-equivalent age: a diffusion tensor and probabilistic tractography study. AJNR Am J Neuroradiol 2011; 32: 20116.
  • 18
    de Bruïne FT, van Wezel-Meijler G, Leijser LM, et al. Tractography of developing white matter of the internal capsule and corpus callosum in very preterm infants. Eur Radiol 2011; 21: 53847.
  • 19
    Thompson DK, Inder TE, Faggian N, et al. Characterization of the corpus callosum in very preterm and full-term infants utilizing MRI. Neuroimage 2011; 55: 47990.
  • 20
    Hasegawa T, Yamada K, Morimoto M, et al. Development of corpus callosum in preterm infants is affected by the prematurity: in vivo assessment of diffusion tensor imaging at term-equivalent age. Pediatr Res 2011; 69: 24954.
  • 21
    Bassi L, Chew A, Merchant N, et al. Diffusion tensor imaging in preterm infants with punctate white matter lesions. Pediatr Res 2011; 69: 5616.
  • 22
    van Kooij BJ, van Pul C, Benders MJ, van Haastert IC, de Vries LS, Groenendaal F. Fibre tracking at term displays sex differences regarding cognitive and motor outcome at 2 years of age in preterm infants. Pediatr Res 2011; 70: 62632.
  • 23
    van Pul C, van Kooij BJ, de Vries LS, Benders MJ, Vilanova A, Groenendaal F. Quantitative fibre tracking in the corpus callosum and internal capsule reveals microstructural abnormalities in preterm infants at term-equivalent age. AJNR Am J Neuroradiol 2012; 33: 67884.
  • 24
    Liu Y, Aeby A, Balériaux D, et al. White matter abnormalities are related to microstructural changes in preterm neonates at term-equivalent age: a diffusion tensor imaging and probabilistic tractography study. AJNR Am J Neuroradiol 2012; 33: 83945.
  • 25
    Ball G, Boardman JP, Aljabar P, et al. The influence of preterm birth on the developing thalamocortical connectome. Cortex 2013; 49: 171121.
  • 26
    Thompson DK, Inder TE, Faggian N, et al. Corpus callosum alterations in very preterm infants: perinatal correlates and 2 year neurodevelopmental outcomes. Neuroimage 2012; 59: 357181.
  • 27
    De Bruïne FT, Van Wezel-Meijler G, Leijser LM, et al. Tractography of white-matter tracts in very preterm infants: a 2-year follow-up study. Dev Med Child Neurol 2013; 55: 42733.
  • 28
    Groppo M, Ricci D, Bassi L, et al. Development of the optic radiations and visual function after premature birth. Cortex 2012; doi:10.1016/j.cortex.2012.02.008. (Epub ahead of print).
  • 29
    Woodward LJ, Anderson PJ, Austin NC, Howard K, Inder TE. Neonatal MRI to predict neurodevelopmental outcomes in preterm infants. N Engl J Med 2006; 355: 68594.
  • 30
    Leijser LM, de Bruïne FT, van der Grond J, Steggerda SJ, Walther FJ, van Wezel-Meijler G. Is sequential cranial ultrasound reliable for detection of white matter injury in very preterm infants? Neuroradiology 2010; 52: 397406.
  • 31
    Bayley N. Bayley Scales of Infant and Toddler Development. San Antonio, TX: PsychCorp, 2006.
  • 32
    Inder TE, Wells SJ, Mogridge NB, Spencer C, Volpe JJ. Defining the nature of the cerebral abnormalities in the premature infant: a qualitative magnetic resonance imaging study. J Pediatr 2003; 143: 1719.
  • 33
    Bayley N, Reuner G. Bayley Scales of Infant Development – Second Edition. Frankfurt: Pearson Assessment & Information, 1993.
  • 34
    Makris N, Kennedy DN, McInerney S, et al. Segmentation of subcomponents within the superior longitudinal fascicle in humans: a quantitative, in vivo, DT-MRI study. Cereb Cortex 2005; 15: 85469.
  • 35
    Iturria-Medina Y, Sotero RC, Canales-Rodríguez EJ, Alemán-Gómez Y, Melie-García L. Studying the human brain anatomical network via diffusion-weighted MRI and Graph Theory. NeuroImage 2008; 40: 106476.
  • 36
    Dubois J, Dehaene-Lambertz G, Perrin M, et al. Asynchrony of the early maturation of white matter bundles in healthy infants: quantitative landmarks revealed noninvasively by diffusion tensor imaging. Hum Brain Mapp 2008; 29: 1427.
  • 37
    Hintz SR, Kendrick DE, Vohr BR, Kenneth Poole W, Higgins RD, Nichd Neonatal Research Network. Sex differences in neurodevelopmental outcomes among extremely preterm, extremely-low-birthweight infants. Acta Paediatr 2006; 95: 123948.
  • 38
    Miller F. Etiology, epidemiology, pathology, and diagnosis. Cerebral Palsy (Chapter: 2). New York: Springer, 2005: 2750.
  • 39
    Hagmann P, Grant PE, Fair DA. MR connectomics: a conceptual framework for studying the developing brain. Front Syst Neurosci 2012; 6: 43.
  • 40
    Burger M, Louw QA. The predictive validity of general movements – a systematic review. Eur J Paediatr Neurol 2009; 13: 40820.
  • 41
    Pannek K, Guzzetta A, Colditz PB, Rose SE. Diffusion MRI of the neonate brain: acquisition, processing and analysis techniques. Pediatr Radiol 2012; 42: 116982.
  • 42
    Tournier JD, Mori S, Leemans A. Diffusion tensor imaging and beyond. Magn Reson Med 2011; 65: 153256.
  • 43
    Jones DK. Studying connections in the living human brain with diffusion MRI. Cortex 2008; 44: 93652.
  • 44
    Wheeler-Kingshott CA, Cercignani M. About ‘axial’ and ‘radial’ diffusivities. Magn Reson Med 2009; 61: 125560.
  • 45
    Behrens TE, Berg HJ, Jbabdi S, Rushworth MF, Woolrich MW. Probabilistic diffusion tractography with multiple fibre orientations: what can we gain? Neuroimage 2007; 34: 14455.
  • 46
    Raffelt D, Tournier JD, Rose S, et al. Apparent fibre density: a novel measure for the analysis of diffusion-weighted magnetic resonance images. Neuroimage 2012; 59: 397694.
  • 47
    Colby JB, Soderberg L, Lebel C, Dinov ID, Thompson PM, Sowell ER. Along-tract statistics allow for enhanced tractography analysis. Neuroimage 2012; 59: 322742.
  • 48
    Deoni SC, Dean DC, O'Muircheartaigh J, Dirks H, Jerskey BA. Investigating white matter development in infancy and early childhood using myelin water faction and relaxation time mapping. Neuroimage 2012; 63: 103853.
  • 49
    Wang Z, Fernández-Seara M, Alsop DC, et al. Assessment of functional development in normal infant brain using arterial spin labelled perfusion MRI. Neuroimage 2008; 39: 9738.
  • 50
    Spittle AJ, Doyle LW, Boyd RN. A systematic review of the clinimetric properties of neuromotor assessments for preterm infants during the first year of life. Dev Med Child Neurol 2008; 50: 25466.
  • 51
    Noble Y, Boyd R. Neonatal assessments for the preterm infant up to 4 months corrected age: a systematic review. Dev Med Child Neurol 2012; 54: 12939.

Supporting Information

  1. Top of page
  2. Abstract
  3. Method
  4. Results
  5. Discussion
  6. Conclusion
  7. Acknowledgements
  8. References
  9. Supporting Information
FilenameFormatSizeDescription
dmcn12250-sup-0001-TableS1.docxWord document91KTable SI: Details of MRI acquisition of included studies.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.