Functional MRI of the newborn

Seminars in Fetal & Neonatal Medicine (2006) 11, 479e488

a v a i l a b l e a t w w w. s c i e n c e d i r e c t . c o m

j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / s i n y

Functional MRI of the newborn Mohamed L. Seghier a,b,*, Francois Lazeyras b, Petra S. Huppi c a

Wellcome Department of Imaging Neuroscience, Institute of Neurology, UCL, 12 Queen Square, London WC1N 3BG, UK b Department of Radiology, Geneva University Hospitals, Micheli-du-Crest 24, 1211 Geneva, Switzerland c Department of Pediatrics, Children’s Hospital of Geneva, 6 rue Willy-Donze´, 1211 Geneva, Switzerland

KEYWORDS Functional MRI; BOLD response; Newborn; Brain activation; Maturation; Brain plasticity

Summary In order to provide accurate prognosis and developmental intervention to newborns, new methods of assessing cerebral functions are needed. The non-invasive technique of functional magnetic resonance imaging (fMRI) can be considered as the leading technique for functional exploration of the infant’s brain. Several studies have previously applied fMRI in both healthy and diseased newborns with different sensory and cognitive tasks. In this chapter, the methodological issues that are proper to the use of fMRI in the newborn are detailed. In addition, an overview of the major findings of previous fMRI studies is provided, with a focus on notable differences from those in adult subjects. More specifically, the functional responses and the localization of cortical activations in healthy and diseased newborns are discussed. We expect a rapid expansion of this field and the establishment of fMRI as a valid clinical diagnostic tool in the newborn. ª 2006 Elsevier Ltd. All rights reserved.

Introduction Brain lesions at birth are often difficult to link to long-term disabilities such as motor, sensory, and cognitive impairments during childhood. With the advent of magnetic resonance imaging (MRI), it has become possible to assess brain lesions and their extent at birth with great precision. Nevertheless, the functional deficits associated with these lesions are extremely difficult to foresee, and there is a lack of predictive evidence of lesions as revealed by structural MRI with actual developmental abilities. In order * Corresponding author. Wellcome Department of Imaging Neuroscience, Institute of Neurology, UCL, 12 Queen Square, London WC1N 3BG, UK. Tel.: þ44 20 7833 7479; fax: þ44 20 7813 1420. E-mail address: [email protected] (Mohamed L. Seghier).

to provide accurate prognosis and developmental intervention to infants that suffer from brain lesions, new methods of assessing cerebral functions are needed. Functional neuroimaging techniques permit the investigation of the functional organization of infants’ and children’s brains. Although functional neuroimaging has produced a wealth of new information in adults, it has been less frequently applied to the developing brain. More specifically, the combination of structural and functional developmental neuroimaging can address the question of where and when maturational changes occur in the brain and how these changes relate to the development of neurosensory, neuromotor, and cognitive abilities.1 As a relatively young field, there are many aspects of human brain development that still need to be explored, such as neurosensory and neuromotor development, development of cognition with learning processes, memory and language acquisition, as well

1744-165X/$ - see front matter ª 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.siny.2006.07.007

480 as brain repair and plasticity after perinatal injury. In this context, functional neuroimaging techniques can bring valuable data in both healthy and diseased newborns.

Functional neuroimaging Several methods for functional exploration of the infant’s brain are accessible.2 These include electroencephalography (EEG), magnetoencephalography (MEG), positron emission tomography (PET), single photon-emission computed tomography (SPECT), magnetic resonance spectroscopy (MRS), functional magnetic resonance imaging (fMRI), and optical topography (OT). Each method is characterized by different methodological and technical parameters and has been used in particular conditions according to its spatial resolution, temporal resolution, degree of invasiveness, sensitivity, and specificity.3,4 For instance, EEG has been used to monitor brain activity to detect seizures after insults but has poor localization power.5 MEG has particularly been used in newborns to functionally assess the maturation of the auditory system.6 With PET, important clues to the metabolic basis of hypoxiceischaemic injury could be provided,7 but the technique is invasive and poorly suited to studying normal development. In addition, the use of SPECT has contributed to the assessment of the relationships between blood flow and metabolic rate changes after neonatal infarction.8 Functional imaging with OT has been shown to be applicable in monitoring brain haemodynamic changes during brain activity of the visual system, but again its applicability to other cognitive functions is limited.9 Although all these methods have been frequently used in infants, fMRI can be considered as the leading technique for functional exploration of the infant’s brain.

Functional MRI fMRI technically refers to the use of the non-invasive MR technology to detect regional changes in signals that are correlated with brain functional activity. This method is widely available and can be performed with any clinical or research MRI scanner with a magnetic field strength of 1.5 T or higher. It is a non-invasive technique that can be used repeatedly and safely on infants, with high temporal and spatial resolutions.10 It can be used serially from birth to school age and adolescence, providing a unique means of studying normal and abnormal development of the brain. Furthermore, MRI techniques appear to be best suited to offering both functional information and high-resolution anatomical details within the same session without exposure to radiation. Principally, fMRI uses deoxygenated haemoglobin (dHb) as an endogenous contrast agent to indirectly depict cortical activated regions. The paramagnetic properties of dHb disrupt the homogeneity of the magnetic fields generated by the powerful permanent magnetic field within the scanner. Generally, neuronal activation within a given cortical region leads to a local increase in the cerebral blood flow (CBF) accompanied by a weak oxygen consumption and therefore to a decrease of dHb concentration in this region. This local dHb concentration drop provokes an alteration of the local magnetic field in the cortical region that fMRI can detect with appropriate sequences. This

M.L. Seghier et al. effect, known as the BOLD (blood-oxygenation-level-dependent) contrast, exploits these haemodynamic modifications that accompany neuronal activity.11 Typically, an fMRI experiment consists of collecting rapid T2*-weighted images with echo planar imaging (EPI) when the subject is executing or submitted to a given task according to a predefined paradigm. The image collected (also referred to as volume or scan) usually contains several slices, and both the number of slices and their thickness are adjusted to obtain whole brain coverage within a time window called the repetition time (TR). Each slice is sampled into spatial units called voxels, and the size of voxels defines the spatial resolution of the experiment. Therefore, fMRI data are generally described as a set of voxels, each with associated time series. fMRI with BOLD contrast can at present determine only relative signal changes between different states of activity. Accordingly, two kinds of paradigm are usually used in fMRI experiments: (1) a block paradigm that alternates between activation and control conditions of tens of seconds in duration; and (2) an event-related paradigm that uses brief presentation of stimuli in a periodic or random manner. These paradigms can be repeated several times in different sessions. The images collected are then reconstructed, pre-processed, and statistically analysed. Several software analyses packages are available, including SPM,12 AFNI,13 FSL,14 MEDx,15 VoxBo,16 and BrainVoyager.17 An overview of the different analysis steps is beyond the scope of this review, but can be found elsewhere.18 Briefly, the first important step is to register and align all images to a reference image (e.g. the first acquired image) by minimizing the global differences between images caused by head motion. Then, aligned images are co-registered and normalized according to a reference template, when group analyses or group comparisons are planned. This step is performed to take into account anatomical variability between subjects during group analyses. The pre-processed images are then submitted to statistical comparison that commonly assumes linear relationships between the paradigm and the detected MRI signal. In this step, the paradigm is expressed in terms of basis functions (e.g. the canonical haemodynamic response function) and compared to the collected time series. Accordingly, each voxel is characterized by a statistical measure (e.g. correlation coefficient, z-score, t-value). Voxels that survive the user-defined threshold are considered as active. These activated voxels are grouped into regions (clusters) and then projected on anatomical images for identification and labelling. More details about methodological and technical issues of fMRI can be found elsewhere.19,20 The non-invasive character of fMRI therefore makes it the neuroimaging technique of choice for studying functional brain development.

fMRI in the fetal period In 1999, the feasibility of fMRI during the fetal period was demonstrated.21 The major motivation behind the use of fMRI to depict fetal activity is the possibility of assessing sensory functions in the fetus, particularly pertinent for obstetricians and paediatricians in determining fetal wellbeing. Although this field is still in its infancy, several studies have successfully assessed fetal brain activity during

Functional MRI of the newborn

481

different tasks, including auditory,21,22 visual,23 and vibroacoustic stimulations.24 Table 1 lists the major characteristics and findings of these studies. Generally, fetal fMRI has been performed at low magnetic fields (0.5 T) by means of tailored abdominal EPI acquisitions with very low spatial resolution (e.g. slice thickness >1 cm), and the BOLD responses detected during the fetal period were positive. This growing field is still facing some methodological problems, in particular motion and signal susceptibility artefacts during image collection.22 A recent review summarizes the advances and difficulties in fetal fMRI.25

fMRI in the newborn Functional MRI has opened up the opportunity to study brain function non-invasively in the newborn. In 1995, Toft et al.26 have used the fMRI technique to visualize changes in cerebral blood flow in infants during ventilator-induced hyperventilation. For the first time in 1996, Born et al.27 showed brain activation in healthy newborns using visual stimulation. To date, more than 20 studies of the development and organization of different brain functions have been successfully performed in the newborn with fMRI.26e48 These studies have principally explored the visual system of healthy newborns, but others have attempted to monitor brain activity during auditory,28,29 sensoryemotor,35,36,45 and language stimulations.34,45 Generally, these studies have been conducted at a magnetic field strength of 1.5 T, with the exception of two studies at 2 T.39,40 They have usually used block paradigm designs, but eventrelated designs have been successfully employed to determine the exact characteristics of the infant BOLD responses.43 In addition, the combination of fMRI with other techniques has been proposed in different studies, including MRI perfusion to assess autoregulation during brain activity,33 diffusion tensor imaging (DTI) to identify the anatomical correlates of brain functions,43,44 structural MRI with T1-weighted images for myelination quantification,48 MRI with T2*-weighted images for cerebral oxidative metabolism assessment,39,42 and optical topography to quantify haemodynamic changes that sustain brain activity.37 Although it is fundamentally equivalent to fMRI in adults, several specific issues should be taken into account when using fMRI in the newborn.49e57 These concern image acquisition, infant monitoring, data processing, and interpretation of results. For image acquisition, using long echo Table 1

times (TE) in EPI sequences is recommended as the T2* of the newborn has been shown to be higher than that of an adult.58 The use of MR-compatible incubators that integrate optimized coils is valuable to maximize safety and signal quality.59 Surface coils can also be employed to improve MRI signal in cortical regions of interest, as shown recently with the occipital lobe.39 In addition, monitoring physiological parameters of the newborn is required during fMRI experiments, including respiratory rate, heart rate, oxygen saturation, blood pressure, and temperature. Although fMRI is probably feasible in awake newborns,29,34 sedation with different anaesthetics is commonly used in newborn subjects (see Table 2). Sedation is generally used to minimize head motion during scanning that could compromise image quality; however, it could limit the choice of tasks and the interpretation of BOLD signal changes. Stimulation should be delivered with adequate devices. For instance, goggles or mirrors fastened to the head coil are used for visual stimulation; compatible headphones or pneumatic earphones, incorporated with standard neonatal ear shields to protect the newborn against scanner noise, are commonly used for auditory stimulation; specific materials that avoid allergic reactions are used for somato-sensorial stimulations. In addition, stimulation characteristics should be adapted according to previous psychophysical studies that have shown differences between infants and adults, including specific frequencies for visual39,43 and auditory29 stimuli. For instance, low frequency for visual stimulation has been used in infants, leading to strong visual cortex activation at, for example, 4 Hz,39 2 Hz,43 and even 1 Hz.29 Some specific issues should also be taken into account during functional data pre-processing: for instance, the small brain size, maturation, grey matter density, and cortex thickness have direct repercussions on the choice of the head motion correction algorithm, kernel size of the spatial smoothing, and the template for spatial normalization in group analysis.53,54 In addition, one major issue in fMRI analysis is the choice of the haemodynamic response function (HRF) used to model the predicted responses that are then statistically compared to the measured time series. Debate still exists about the exact nature, shape, timing parameters, and linearity of HRF in newborns.34,43,54 The complexity of HRF identification here is related to several physiological parameters, including the particular architecture of the capillary bed, vessel orientation, synaptic density, and maturation.49 Statistical thresholds should be adjusted

List of functional magnetic resonance imaging (fMRI) studies performed during the fetal period

Study

Subjects

Gestational age (weeks)

Stimuli

MRI scanner resolution (mm3)

BOLD responses Frontal lobe (5), no activation (3), artefacts (1) Temporal lobe (7), no activation (8), artefacts (2) Temporal lobe (2), no activation (1), artefacts (1) Temporal lobe (4), Frontal lobe (1), no activation (2), artefacts (5)

Fulford et al., 200323

9

>36

Visual

0.5 T (4  4  10)

Fulford et al., 200424

17

>36

Vibroacoustic

0.5 T (4  4  10)

38e39.4

Auditory

0.5 T (5  5  15)

>37

Auditory

0.5 T (5  5  15)

Hykin et al., 199921

4

Moore et al. 200122

12

BOLD, blood-oxygenation-level-dependent.

482

Table 2

List of functional magnetic resonance imaging (fMRI) studies with healthy and diseased newborns

Study

Subjects

Age

Sedation

Stimuli

Healthy newborns Altman and Bernal, 200128

38 (40 total)

2 months to 9 years

Visual

MRI scanner

Combination

BOLD responses

1.5 T

e

Visual: positive (18), negative (28) Auditory: positive (all)

Auditory Visual Visual Visual

1.5 T 1.5 T 1.5 T 1.5 T

e e e e

Positive (5), negative (9) Negative Negative Positive (4), negative (3)

Visual

1.5 T

Perfusion assessment with MRI (FAIR) e

Negative

Motor: positive (6), negative (5) Visual: positive (all) Both negative and positive (depends on responses’ laterality) In V1: positive (age <60 days), negative (age >60 days) In LGN: positive (all) Both negative and positive

14 (20 total) 7 17 8 (37 total)

1 6 3 7

Born et al., 200233

4

4e71 months

Dehaene-Lambertz et al., 200234 Erberich et al., 200335

20 (30 total)

8e14 weeks

None

Language

1.5 T

7 preterm

24e39 weeks

Chloral hydrate

1.5 T, incubator

e

Erberich et al., 200636

24 (42 total)

e

Chloral hydrate

Sensorye motor Visual Sensorye motor

1.5 T, incubator

e

Konishi et al., 200237

27

1 day to 22 weeks

Pentobarbital

Visual

Marcar et al., 200439

21

3 months to 12 years

Midazolam (16), none (5)

Visual

2 T, surface coil

Martin et al., 199940

58 (75 total)

1 day to 12 years

Visual

2T

Morita et al., 200041

16 (26 total)

1e12 months

Chloral hydrate (12), pentobarbital (28), halothane/N2O (9), none (9) Pentobarbital

Visual

1.5 T

e

Muramoto et al., 200242

20 (42 total)

1 day to 32 weeks

Pentobarbital

Visual

1.5 T

e

Promethazine, pethidine, droperidol, diazepam

Visual

1.5 T

Cerebral metabolite rate of oxygen estimated by MRI e

Sie et al., 200146

7 (42 total)

day to 11 weeks weeks to 36 months days to 48 months days to 7 months

Auditory

1.5 T

Deoxy-haemoglobin assessment by optical topography Cerebral metabolite rate of oxygen estimated by MRI e

Positive

Positive (23), negative (25)

In V1: positive (age <60 days), negative (age >60 days) In LGN: positive (all) Positive (age <8 weeks), negative (age >8 weeks) Negative (4)

M.L. Seghier et al.

Anderson et al., 200129 Born et al., 199627 Born et al. 199831 Born et al., 200032

Chloral hydrate (9), pentobarbital (25), propofol (1), alprazolam (3) None Chloral hydrate Chloral hydrate Chloral hydrate (5), none (2) Chloral hydrate

8 term, 7 preterm

1e54 weeks

Pentobarbital

Visual

1.5 T

e

Yamada et al., 200048

27

1 day to 22 weeks

Pentobarbital

Visual

1.5 T

Myelination quantification by MRI

Positive (age <8 weeks), negative (age >8 weeks) Positive (age <7 weeks), negative (age >8 weeks)

11 months

Chloral hydrate

Visual

1.5 T

e

Negative

4 days to 6 years

Visual

1.5 T

e

Visual Visual

1.5 T 1.5 T

Positive (2), negative (16) Negative Negative

Diseased newborns Bernal and Altman, 200430 Born et al., 200032

18 (37 total)

Liu et al., 200038 Seghier et al., 200443

1

36 months 3 months

Chloral hydrate (16), none (3) Nembutal Chloral hydrate

Seghier et al., 200544

1

20 months

Pentobarbital

Visual

1.5 T

Sie et al., 200146

21 (42 total)

18 months

Visual

1.5 T

Souweidane et al., 199945

8

3 months to 8 years

Promethazine, pethidine, droperidol, diazepam Propofol

e White matter fibres tracking with diffusion imaging (DTI) White matter fibres tracking with diffusion imaging (DTI) e

Visual Tactile Language

1.5 T

e

2 (3 total)

1 (2 total)

Functional MRI of the newborn

Yamada et al., 199747

Negative

Positive (1), negative (13)

Positive

FAIR, flow-sensitive alternating inversion recovery; BOLD, blood-oxygenation-level-dependent; V1, primary visual area; LGN, lateral geniculate nucleus.

483

484

M.L. Seghier et al.

differently from the standards in adults, and less conservative thresholds have been previously suggested in fMRI studies of the newborn (e.g. the BOLD signal might be smaller in infants and young children than in adults39). Variable thresholds across subjects have also been employed in some studies with infants.32 In this context, absence of activation in certain brain regions should be carefully interpreted as this does not necessarily mean that these regions are not participating in the task.51 Finally, cortical representations in newborns could potentially be different from those in adults and therefore should be taken into account during interpretation of results: for example, a more anterior localization of visual activation and weak laterality for somato-sensorial areas. These different practical issues are summarized in Fig. 1.

fMRI to assess functional brain development in newborns The characteristics of previous fMRI studies of the newborn are listed in Table 2. These studies have implicated variable numbers of infants at different ages, used different sedation protocols, and employed different tasks. About 90% of these studies have used visual stimuli. Their major findings concern the site of visual activation and the sign of the BOLD responses. In general, activated foci were located within the medial occipital cortex in more anterior regions

Figure 1 newborn.

as compared to adults. This anterior localization has been attributed to different factors, including the absence of spatial contrast in the visual stimulation used and the stimulation of more peripheral areas via the closed eyes of the infants.28,31,41 This occipital region has been identified as the primary visual area, V1, but others have alternatively considered it as the secondary visual area, V2.40 Moreover, the detected BOLD responses were negative, which is different from the case in adult subjects. The sign of the BOLD response may depend on the age of the infant, with positive responses in newborns under 2 months of age and negative responses in newborns over 2 months of age.37,41,47,48 Different hypotheses have been put forward. One hypothesis is that infants are usually sleeping during the fMRI experiment, which could lead to negative responses.60 An alternative explanation can be related to the large increase in oxygen consumption during visual stimulation due to rapid synaptogenesis at this early stage of life.42 Interestingly, the lateral geniculate nucleus (LGN)37,41,42 and the thalamic pulvinar40,51 have been reported in some studies, with generally positive BOLD responses unrelated to the age of the infants. Concerning the auditory system, two studies have used auditory stimuli with block designs.28,29 In 14 non-sedated infants, significant responses have been detected bilaterally in the superior temporal regions similar to the adult auditory system.29 More specifically, nine infants have presented negative responses even without the use of any

Major methodological issues that should be taken into account in functional magnetic resonance imaging (fMRI) of the

Functional MRI of the newborn sedative agent. In another study, 26 infants out of a total of 38 have presented activations in different bilateral frontal and temporo-parietal regions, including the primary auditory cortex, the secondary auditory cortex, and the frontal lobe.28 All the responses were positive in these sedated infants. The differences between these two studies are principally related to the use of different auditory stimuli (tone29 and speech28). Furthermore, the language system has also been explored with fMRI. In 20 healthy non-sedated infants, aging 2e3 months, speech perception was explored in a block paradigm.34 Activation maps include prefrontal areas as well as temporo-parietal network, which are typically activated with language paradigms in adults. Interestingly, positive responses were observed in the angular gyrus and the precuneus when normal speech was compared to reversed speech, and these activations were significantly dominated by the left hemisphere.34 These findings might suggest that the cortical substrates of the human language are formed before the onset of speech production. The sensoryemotor system has recently been investigated in seven preterm infants.35 Only two infants have shown responses equivalent to the primary sensoryemotor cortex similar to those shown in adults, whereas the remaining infants have shown activations in a medial region probably equivalent to the supplementary motor area and in some anterior frontal regions. More recently, in 24 sedated newborns, fMRI with somato-sensorial stimulation has shown weak hemispheric dominance for somato-sensory areas,36 but a trend towards lateralization was observed in older infants. These different fMRI studies of the healthy newborn have shown interesting results about the functional subdivisions of the infant’s brain and have underlined some notable differences from the adult’s brain. However, it is worth noting that the percentage of failures (i.e. BOLD responses not detected in some subjects) is higher than in older subjects (about 30% of subjects have presented no functional activations).34,40 This is generally due to different artefacts (e.g. considerable head motion) or haemodynamic modifications (e.g. alteration of both CBF and oxygen metabolism). This failure percentage, however, does not constitute a major handicap for fMRI of the newborn, and similar percentages have been also observed in fMRI of young children.61

fMRI in newborn brain pathology Relevant information may be obtained by fMRI when standard clinical investigations are not conclusive, providing diagnosis and guiding therapeutic interventions. Several interesting studies that support the utility of fMRI as a valid diagnostic tool have investigated functional responses after brain insult, particularly in infants with an abnormal visual system.30,32,38,43e46 Table 2 presents the major technical characteristics of these studies. Previously, in a 36-month-old infant having holoprosencephaly with severely malformed occipital lobes, fMRI with visual stimulation has shown significant negative responses in the posterioremedial portions of both hemispheres,

485 including striate cortex and visual secondary areas.38 These results are important in the planning of epilepsy surgery in order to avoid potential damage to the eloquent visual cortex. In addition, fMRI has been shown to be the unique tool to map visual function in patients with SturgeeWeber syndrome (encephalotrigeminal angiomatosis). In two female infants of 11 months of age, visual stimulation has shown that the SturgeeWeber syndrome yields abnormal (e.g. asymmetric or absent) activations of the visual system in comparison with healthy control subjects.30 The assessment of the viability of the visual cortex is important to provide guidance in regard to surgical planning and may be helpful to predict post-surgical deficits. In addition, in 21 infants with periventricular leukomalacia, significant activations have been observed in the anterior part of the primary visual cortex with variable extension to secondary visual areas.46 Although the visual responses were comparable to control healthy infants, the activated cortical volume in the visual cortex of the injured infants has been shown to be highly correlated with the degree of severity of the occipital periventricular leukomalacia. Moreover, the feasibility of fMRI with visual, tactile, and language stimuli has been shown in infants of 33 months of age having different cerebral pathologies.45 This study has documented the reliability of several passive stimulations in sedated infants. Recently, we reported an infant with structural MRI in the newborn period showing a large lesion involving the left temporo-parieto-occipital region, compatible with acute stage of infarction. During early childhood, clinical assessment of the visual field is difficult, and unilateral central blindness is therefore impossible to diagnose. fMRI was performed to assess functional outcome of this unilateral perinatal lesion (Fig. 2). With visual stimuli presented in both block and event-related paradigms, fMRI showed a negative BOLD activation in the visual cortex of the intact right hemisphere, principally in the anterior part, but no activation in the injured left hemisphere.43 A second fMRI experiment was performed in the same infant at 20 months of age, showing significant negative activation in the visual cortex of the injured left hemisphere.44 This functional recovery occurred in parallel with new structural development that was monitored by DTI techniques. These neuroimaging findings might contribute to the comprehension of mechanisms of plasticity and to the elaboration of a pertinent strategy in terms of diagnosis and rehabilitation in very young infants. In a large number of infants with brain damage and variable neurological outcomes, fMRI with visual stimuli has provided interesting functional information about the visual deficits in these infants.32 More specifically, in two infants with cerebral infarction and cerebrovascular disease, visual activation was unilateral or strongly asymmetrical towards the healthier hemisphere. In addition, in four infants with established visual deficits, fMRI has shown less activation in the visual cortex as compared to healthy infants, suggesting the possibility of detecting functional deficits with fMRI in very young infants. Moreover, with ten infants examined twice at different ages, this study has particularly emphasized the important role of longitudinal fMRI studies to investigate plasticity and recovery of the visual system after brain insult.32

486

M.L. Seghier et al.

Figure 2 Longitudinal functional magnetic resonance imaging (fMRI) study with visual stimulation in an infant with temporoparieto-occipital lesion. Some recovery mechanisms are responsible for the negative responses observed in the injured left hemisphere.

Perspectives The above-mentioned studies demonstrate that fMRI can reveal functional networks in children comparable to those found in adults, opening the way to study brain maturation and learning processes in normal children. This will furthermore permit the study of developmental mechanisms in children with cognitive impairments. Nevertheless, to achieve this goal, one needs to be able to collect normative data in infants and young children without sedation. This will require the development of specific psychological methods to direct the baby’s attention towards adapted stimuli. Motion may be a severe limitation in using MRI techniques in young children. Tailored acquisition with prospective motion correction will be necessary in order to collect sound fMRI data, especially for infants a few months old. Parallel imaging with ultra-fast sequences are also potentially interesting to shorten acquisition duration. In addition, diffusion tensor imaging (DTI) acquisition can be performed in the same session as the fMRI experiment. DTI offers a means to assess cerebral connectivity and can potentially track the development of white matter fibres.

A recent study on normal development reports a positive correlation between DTI parameters and BOLD response in specific brain areas.62 fMRI can further be used to monitor pharmacological treatment effects in young children.63 Moreover, the assessment of different cognitive tasks are potentially relevant in the newborn to understand the origins of developmental delay.55 Previous attempts to study the language system in infants have opened the way for other tasks. For instance, studying attention or emotion, with auditory modality, could help to understand how these affective components are functionally coded in the very young infant. Such information is very suitable to elucidate the mechanisms of learning and cognitive development at this early stage of life. With this objective, longitudinal studies with various cognitive tasks have yet to be performed. The combination of the non-invasive neuroimaging techniques of fMRI and DTI are destined to study functional connectivity that may permit the identification of how cortical regions interact together during a given cognitive task and how plasticity can regain function after perinatal brain injury. For further information on related topics see de Vries,64 Lowery et al.,65 and Huotilainen.66

Functional MRI of the newborn

Research directions  Characterization of the BOLD response in the newborn.  Combination of fMRI with connectivity assessment by DTI.  Longitudinal studies with different cognitive tasks for visualizing brain development and plasticity, in particular learning processes and hemispheric specialization.  Characterization of inter-individual functional variability in the newborn and constitution of normative databases.  Validation and comparison of fMRI findings to standard clinical tools.

References 1. Johnson MH. Functional brain development in humans. Nat Rev Neurosci 2001;2:475e83. 2. Munson S, Schroth E, Ernst M. The role of functional neuroimaging in pediatric brain injury. Pediatrics 2006;117:1372e81. 3. Toga AW, Mazziotta JC. Brain mapping: the methods. London: Academic Press; 1996. p. 471. 4. Ernst M, Rumsey JM. Functional neuroimaging in child psychiatry. Cambridge: Cambridge University Press; 2000. p. 440. 5. Hellstrom-Westas L, Rosen I. Electroencephalography and brain damage in preterm infants. Early Hum Dev 2005;81: 255e61. 6. Holst M, Eswaran H, Lowery C, Murphy P, Norton J, Preissl H. Development of auditory evoked fields in human fetuses and newborns: a longitudinal MEG study. Clin Neurophysiol 2005; 116:1949e55. 7. Chugani HT. Positron emission tomography scanning: applications in newborns. Clin Perinatol 1993;20:395e409. 8. Kusaka T, Ijichi S, Yamamoto Y, Nishiyama Y. Changes in cerebral glucose metabolism in newborn infants with cerebral infarction. Pediatr Neurol 2005;32:46e9. 9. Hebden JC. Advances in optical imaging of the newborn infant brain. Psychophysiology 2003;40:501e10. 10. Menon RS, Kim S-G. Spatial and temporal limits in cognitive neuroimaging with fMRI. Trends Cogn Sci 1999;3:207e16. 11. Ogawa S, Menon RS, Kim SG, Ugurbil K. On the characteristics of functional magnetic resonance imaging of the brain. Annu Rev Biophys Biomol Struct 1998;27:447e74. 12. http://www.fil.ion.ucl.ac.uk/spm/. 13. http://afni.nimh.nih.gov/. 14. http://www.fmrib.ox.ac.uk/. 15. http://medx.sensor.com/. 16. http://www.voxbo.org/. 17. http://www.brainvoyager.com/12. 18. Smith SM. Overview of fMRI analysis. Br J Radiol 2004;77: S167e75. 19. Hennig J, Speck O, Koch MA, Weiller C. Functional magnetic resonance imaging: a review of methodological aspects and clinical applications. J Magn Reson Imaging 2003;18:1e15. 20. Faro SH, Mohamed FB. Functional MRI: basic principles and clinical applications. New York: Springer; 2006. p. 470. 21. Hykin J, Moore R, Duncan K, Clare S, Baker P, Johnson I, et al. Fetal brain activity demonstrated by functional magnetic resonance imaging. Lancet 1999;354:645e6. 22. Moore RJ, Vadeyar S, Fulford J, Tyler DJ, Gribben C, Baker PN, et al. Antenatal determination of fetal brain activity in

487

23.

24.

25. 26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

40.

41.

42.

response to an acoustic stimulus using functional magnetic resonance imaging. Hum Brain Mapp 2001;12:94e9. Fulford J, Vadeyar SH, Dodampahala SH, Moore RJ, Young P, Baker PN, et al. Fetal brain activity in response to a visual stimulus. Hum Brain Mapp 2003;20:239e45. Fulford J, Vadeyar SH, Dodampahala SH, Ong S, Moore RJ, Baker PN, et al. Fetal brain activity and hemodynamic response to a vibroacoustic stimulus. Hum Brain Mapp 2004;22: 116e21. Gowland P, Fulford J. Initial experiences of performing fetal fMRI. Exp Neurol 2004;190:S22e7. Toft PB, Leth H, Lou HC, Pryds O, Peitersen B, Henriksen O. Local vascular CO2 reactivity in the infant brain assessed by functional MRI. Pediatr Radiol 1995;25:420e4. Born P, Rostrup E, Leth H, Peitersen B, Lou HC. Change of visually induced cortical activation patterns during development. Lancet 1996;347:543. Altman NR, Bernal B. Brain activation in sedated children: auditory and visual functional MR imaging. Radiology 2001;221: 56e63. Anderson AW, Marois R, Colson ER, Peterson BS, Duncan CC, Ehrenkranz RA, et al. Neonatal auditory activation detected by functional magnetic resonance imaging. Magn Reson Imaging 2001;19:1e5. Bernal B, Altman N. Visual functional magnetic resonance imaging in patients with SturgeeWeber syndrome. Pediatr Neurol 2004;31:9e15. Born P, Leth H, Miranda MJ, Rostrup E, Stensgaard A, Peitersen B, et al. Visual activation in infants and young children studied by functional magnetic resonance imaging. Pediatr Res 1998;44:578e83. Born AP, Miranda MJ, Rostrup E, Toft PB, Peitersen B, Larsson HB, et al. Functional magnetic resonance imaging of the normal and abnormal visual system in early life. Neuropediatrics 2000;31:24e32. Born AP, Rostrup E, Miranda MJ, Larsson HB, Lou HC. Visual cortex reactivity in sedated children examined with perfusion MRI (FAIR). Magn Reson Imaging 2002;20:199e205. Dehaene-Lambertz G, Dehaene S, Hertz-Pannier L. Functional neuroimaging of speech perception in infants. Science 2002; 298:2013e5. Erberich SG, Friedlich P, Seri I, Nelson Jr MD, Bluml S. Functional MRI in neonates using neonatal head coil and MR compatible incubator. Neuroimage 2003;20:683e92. Erberich SG, Panigrahy A, Friedlich P, Seri I, Nelson MD, Gilles F. Somatosensory lateralization in the newborn brain. NeuroImage 2006;29:155e61. Konishi Y, Taga G, Yamada H, Hirasawa K. Functional brain imaging using fMRI and optical topography in infancy. Sleep Med 2002;3:S41e3. Liu GT, Hunter J, Miki A, Fletcher DW, Brown L, Haselgrove JC. Functional MRI in children with congenital structural abnormalities of the occipital cortex. Neuropediatrics 2000;31: 13e5. Marcar VL, Strassle AE, Loenneker T, Schwarz U, Martin E. The influence of cortical maturation on the BOLD response: an fMRI study of visual cortex in children. Pediatr Res 2004; 56:967e74. Martin E, Joeri P, Loenneker T, Ekatodramis D, Vitacco D, Hennig J, et al. Visual processing in infants and children studied using functional MRI. Pediatr Res 1999;46:135e40. Morita T, Kochiyama T, Yamada H, Konishi Y, Yonekura Y, Matsumura M, et al. Difference in the metabolic response to photic stimulation of the lateral geniculate nucleus and the primary visual cortex of infants: a fMRI study. Neurosci Res 2000;38:63e70. Muramoto S, Yamada H, Sadato N, Kimura H, Konishi Y, Kimura K, et al. Age dependent change in metabolic response

488

43.

44.

45.

46.

47.

48.

49.

50. 51. 52. 53. 54.

M.L. Seghier et al. to photic stimulation of the primary visual cortex in infants: functional magnetic resonance imaging study. J Comput Assist Tomogr 2002;26:894e901. Seghier ML, Lazeyras F, Zimine S, Maier SE, Hanquinet S, Delavelle J, et al. Combination of event-related fMRI and diffusion tensor imaging in an infant with perinatal stroke. NeuroImage 2004;21:463e72. Seghier ML, Lazeyras F, Zimine S, Saudan-Frei S, Safran AB, Huppi PS. Visual recovery after perinatal stroke evidenced by functional and diffusion MRI. BMC Neurol 2005;5:17. Souweidane MM, Kim KH, McDowall R, Ruge MI, Lis E, Krol G, et al. Brain mapping in sedated infants and young children with passive-functional magnetic resonance imaging. Pediatr Neurosurg 1999;30:86e92. Sie LTL, Rombouts SA, Valk IJ, Hart AA, Scheltens P, van der Knaap MS. Functional MRI of visual cortex in sedated 18 month-old infants with or without periventricular leukomalacia. Dev Med Child Neurol 2001;43:486e90. Yamada H, Sadato N, Konishi Y, Kimura K, Tanaka M, Yonekura Y, et al. A rapid brain metabolic change in infants detected by fMRI. Neuroreport 1997;8:3775e8. Yamada H, Sadato N, Konishi Y, Muramoto S, Kimura K, Tanaka M, et al. A milestone for normal development of the infantile brain detected by functional MRI. Neurology 2000; 55:218e23. Gaillard WD, Grandin CB, Xu B. Developmental aspects of pediatric fMRI: considerations for image acquisition, analysis, and interpretation. Neuroimage 2001;13:239e49. Logan W. Functional magnetic resonance imaging in children. Semin Pediatr Neurol 1999;6:78e86. Martin E, Marcar VL. Functional MR imaging in pediatrics. Magn Reson Imaging Clin N Am 2001;9:231e45. Seyffert M, Castellanos FX. Functional MRI in pediatric neurobehavioral disorders. Int Rev Neurobiol 2005;67:239e84. Bookheimer SY. Methodological issues in pediatric neuroimaging. Ment Retard Dev Disabil Res Rev 2000;6:161e5. Poldrack RA, Pare-Blagoev EJ, Grant PE. Pediatric functional magnetic resonance imaging: progress and challenges. Top Magn Reson Imaging 2002;13:61e70.

55. Thomas KM, Casey BJ. Functional MRI in pediatrics. In: Moonen CTW, Bandettini PA, editors. Functional MRI. Springer-Verlag; 1999. p. 513e23. 56. Wilke M, Holland SK, Myseros JS, Schmithorst VJ, Ball WS. Functional magnetic resonance imaging in pediatrics. Neuropediatrics 2003;34:225e33. 57. Davidson MC, Thomas KM, Casey BJ. Imaging the developing brain with fMRI. Ment Retard Dev Disabil Res Rev 2003;9: 161e7. 58. Rivkin MJ, Wolraich D, Als H, McAnulty G, Butler S, Conneman N, et al. Prolonged T*2 values in newborn versus adult brain: implications for fMRI studies of newborns. Magn Reson Med 2004;51:1287e91. 59. Whitby EH, Griffiths PD, Lonneker-Lammers T, Srinivasan R, Connolly DJ, Capener D, et al. Ultrafast magnetic resonance imaging of the neonate in a magnetic resonance-compatible incubator with a built-in coil. Pediatrics 2004;113:e150e2. 60. Born AP, Law I, Lund TE, Rostrup E, Hanson LG, Wildschiødtz G, et al. Cortical deactivation induced by visual stimulation in human slow-wave sleep. Neuroimage 2002;17:1325e35. 61. Byars AW, Holland SK, Strawsburg RH, Bommer W, Dunn RS, Schmithorst VJ, et al. Practical aspects of conducting largescale functional magnetic resonance imaging studies in children. J Child Neurol 2002;17:885e90. 62. Olesen PJ, Nagy Z, Westerberg H, Klingberg T. Combined analysis of DTI and fMRI data reveals a joint maturation of white and grey matter in a fronto-parietal network. Cogn Brain Res 2003;18:48e57. 63. Vaidya CJ, Austin G, Kirkorian G, Ridlehuber HW, Desmond JE, Glover GH, et al. Selective effects of methylphenidate in attention deficit hyperactivity disorder: a functional magnetic resonance study. Proc Natl Acad Sci U S A 1998;95: 14494e9. 64. de Vries JIP. Fetal brain monitoring: future applications. Semin Fetal Neonatal Med 2006;11:423e9. 65. Lowery CL, Eswaran H, Murphy P, Preissl H. Fetal magnetoencephalography. Semin Fetal Neonatal Med 2006;11:430e6. 66. Huotilainen M. Magnetoencephalography of the newborn brain. Semin Fetal Neonatal Med 2006;11:437e43.