Claire H. Salmond

Distributional assumptions in voxel-based morphometry

In this paper we address the assumptions about the distribution of errors made by voxel-based morphometry. Voxel-based morphometry (VBM) uses the general linear model to construct parametric statistical tests. In order for these statistics to be valid, a small number of assumptions must hold. A key assumption is that the models error terms are normally distributed. This is usually ensured through the Central Limit Theorem by smoothing the data. However, there is increasing interest in using minimal smoothing (in order to sensitize the analysis to regional differences at a small spatial scale). The validity of such analyses is investigated. In brief, our results indicate that nonnormality in the error terms can be an issue in VBM. However, in balanced designs, provided the data are smoothed with a 4-mm FWHM kernel, nonnormality is sufficiently attenuated to render the tests valid. Unbalanced designs appear to be less robust to violations of normality: a significant number of false positives arise at a smoothing of 4 and 8 mm when comparing a single subject to a group. This is despite the fact that conventional group comparisons appear to be robust, remaining valid even with no smoothing. The implications of the results for researchers using voxel-based morphometry are discussed.

Bilateral brain abnormalities associated with dominantly inherited verbal and orofacial dyspraxia

The KE family is a large three‐generational pedigree in which half of the members suffer from a verbal and orofacial dyspraxia in association with a point mutation in the FOXP2 gene. This report extends previous voxel‐based morphometric analyses of magnetic resonance imaging (MRI) scans (Watkins et al. [ 2002 ] Brain 125:465–478) using a bilateral conjunction analysis. This searches specifically for areas of grey matter density that differ bilaterally in the affected members compared with both matched controls and the unaffected family members. 3‐D T1‐weighted MRI datasets of 17 family members (10 affected, 7 unaffected) and matched controls were compared. The most significant findings were reduced grey matter density bilaterally in the caudate nucleus, the cerebellum, and the left and right inferior frontal gyrus in the affected members. In addition, increased grey matter density was found bilaterally in the planum temporale. These results confirm that a point mutation in FOXP2 is associated with several bilateral grey matter abnormalities in both motor and language related regions. The results also demonstrate the advantages of using a conjunction analysis when bilateral abnormalities are suspected. Hum. Brain Mapping 18:194–200, 2003.

A direct test for lateralization of language activation using fMRI: comparison with invasive assessments in children with epilepsy.

This study introduces a direct method of assessing cerebral lateralization for language based on fMRI activation. The method, derived from a voxel-based morphometry study by C. H. Salmond et al. (2000, Hum. Brain Mapping 11, 223-232), bases lateralization on the direct statistical comparison of the magnitude of task-induced activation in homotopic regions of the two hemispheres. Lateralization results obtained with this direct method were compared to those obtained with a widely used method which involves the calculation of a laterality index (LI) based on the number of significantly activated voxels in the inferior frontal gyrus of each hemisphere. In order to compare the validity of the two methods, a covert verb-generation task was performed by eight children with epilepsy whose language lateralization was examined using invasive techniques. Lateralization results derived from fMRI activation showed that the calculation of a LI presented some limitations. Importantly, the LI value was dependent on the activation threshold chosen to calculate that LI. As a consequence, the correlation between the LI and the invasive methods could vary with the chosen threshold. By contrast, the proposed direct method gave some indication of the reliability of the lateralization and provided results that, in all eight children, were consistent with those obtained using invasive techniques. It is suggested that the direct method could be used in future fMRI studies to establish hemispheric lateralization for cognitive functions.

Developmental amnesia: Effect of age at injury

Hypoxic–ischemic events sustained within the first year of life can result in developmental amnesia, a disorder characterized by markedly impaired episodic memory and relatively preserved semantic memory, in association with medial temporal pathology that appears to be restricted to the hippocampus. Here we compared children who had hypoxic–ischemic events before 1 year of age (early group, n = 6) with others who showed memory problems after suffering hypoxic–ischemic events between the ages of 6 and 14 years (late group, n = 5). Morphometric analyses of the whole brain revealed that, compared with age-matched controls, both groups had bilateral abnormalities in the hippocampus, putamen, and posterior thalamus, as well as in the right retrosplenial cortex. The two groups also showed similar reductions (≈40%) in hippocampal volumes. Neuropsychologically, the only significant differences between the two were on a few tests of immediate memory, where the early group surpassed the late group. The latter measures provided the only clear indication that very early injury can lead to greater functional sparing than injury acquired later in childhood, due perhaps to the greater plasticity of the infant brain. On measures of long-term memory, by contrast, the two groups had highly similar profiles, both showing roughly equivalent preservation of semantic memory combined with marked impairment in episodic memory. It thus appears that, if this selective memory disorder is a special syndrome related to the early occurrence of hypoxia-induced damage, then the effective age at injury for this syndrome extends from birth to puberty.

The role of the medial temporal lobe in autistic spectrum disorders

The neural basis of autistic spectrum disorders (ASDs) is poorly understood. Studies of mnemonic function in ASD suggest a profile of impaired episodic memory with relative preservation of semantic memory (at least in high‐functioning individuals). Such a pattern is consistent with developmental hippocampal abnormality. However, imaging evidence for abnormality of the hippocampal formation in ASD is inconsistent. These inconsistencies led us to examine the memory profile of children with ASD and the relationship to structural abnormalities. A cohort of high‐functioning individuals with ASD and matched controls completed a comprehensive neuropsychological memory battery and underwent magnetic resonance imaging for the purpose of voxel‐based morphometric analyses. Correlations between cognitive/behavioural test scores and quantified results of brain scans were also carried out to further examine the role of the medial temporal lobe in ASD. A selective deficit in episodic memory with relative preservation of semantic memory was found. Voxel‐based morphometry revealed bilateral abnormalities in several areas implicated in ASD including the hippocampal formation. A significant correlation was found between parental ratings reflecting autistic symptomatology and the measure of grey matter density in the junction area involving the amygdala, hippocampus and entorhinal cortex. The data reveal a pattern of impaired and relatively preserved mnemonic function that is consistent with a hippocampal abnormality of developmental origin. The structural imaging data highlight abnormalities in several brain regions previously implicated in ASD, including the medial temporal lobes.

Heterogeneity in the Patterns of Neural Abnormality in Autistic Spectrum Disorders: Evidence from ERP and MRI

Autistic spectrum disorder (ASD) refers to a heterogeneous group of social communication problems. Research into the neural basis of ASD has revealed abnormalities in a number of different regions of the brain. However, the literature is highly inconsistent. One possible explanation for these discrepancies is differences in intelligence. Children with ASD and below average intelligence may be hypothesised to show additional or different neural abnormalities. This possibility was explored using structural magnetic resonance imaging (MRI) and event-related potentials (ERP). Two groups of children with ASD were studied, those with average or above average intelligence (high ASD group) and those with below average intelligence (low ASD group). The structural MRI data were analysed using voxel-based morphometry (VBM). Using the family-wise error threshold, results showed bilateral abnormality common to the two ASD groups in the cerebellum, fusiform gyrus and frontal cortex. In addition, a number of regions were found to be significantly different in the two ASD groups: regions of the cerebellum showed increased grey matter density bilaterally in the high ASD group, but decreased grey matter density bilaterally in the low ASD group. Further, compared to the high ASD group, additional bilateral abnormalities were found in the postcentral gyrus and regions of the dorsolateral prefrontal cortex in the low ASD group. Using the less stringent false discovery rate (FDR) threshold, differences were also seen in the medial temporal lobes. ERPs also showed differences between the two ASD groups. Whereas the ERPs of the high ASD group were not significantly different from those of the controls, the low ASD group had delayed novelty P3a responses and reduced amplitude target P3b components. These data provide convergent ERP and MRI evidence for the heterogeneity of neural abnormality in ASD in relation to variations in intelligence.

Detecting bilateral abnormalities with voxel-based morphometry.

In this article we describe a new method, using SPM99, that searches explicitly for bilateral structural abnormalities. Children with bilateral pathology have a poorer prognosis than children with unilateral damage. After brain injury or disease in childhood, it is thought that rescue of function is only possible if the neuronal substrates of that function are preserved and operational in at least one hemisphere [Vargha‐Khadem and Mishkin, 1997 ]. If this is the case, the detection of bilateral abnormalities would greatly facilitate more accurate prognosis in children with brain injury or developmental disorders. We have therefore developed a technique to detect bilateral abnormalities that uses conjunction analysis with voxel based morphometry. It is illustrated using a group of patients with bilateral hypoxic‐ischaemic damage to the hippocampus. The approach is shown to have enhanced specificity and sensitivity relative to conventional unilateral characterisations. Hum. Brain Mapping 11:223–232, 2000.

Memory in Autism: A comparison of memory profiles in relation to neuropathology in autism, developmental amnesia and children born prematurely

© Cambridge University Press 2008 and Cambridge University Press, 2009. Introduction Autistic spectrum disorders (ASDs) are characterized by impairments in social interaction and communication, and by restricted or repetitive behaviours and interests. The degree of impairment varies enormously. For example, individuals with low-functioning autism (LFA) may be mute or have atypical language, whilst individuals with high-functioning autism (HFA – used here to include individuals with Asperger syndrome) may have good language but nevertheless suffer from communication problems including impaired nonverbal communication. The underlying neural abnormalities associated with autism-related disorders are not well understood. In addition to investigations of the defining behavioural impairments, as listed above, a number of studies have examined memory functions in people with ASDs. These have revealed an impairment in episodic memory (i.e. memory for events and episodes; Russell 1996; Bowler, Gardiner & Grice, 2000; Millward et al., 2000; Gardiner, Bowler & Grice, 2003), but not semantic memory (i.e. memory for facts or world knowledge – Tulving, 1972), at least in higher-functioning individuals (e.g. Ameli et al., 1988; Minshew et al., 1992; Bennetto, Pennington & Rogers, 1996; Siegel, Minshew & Goldstein, 1996; Farrant, Blades & Boucher, 1998). Previous reports of individuals with developmental amnesia (DA) and a group of children born preterm (PT) have revealed a similar dissociation between episodic memory, which is selectively impaired, and factual memory, which is relatively preserved (e.g. Vargha-Khadem et al., 1997, 2003; Gadian et al., 2000; Isaacs et al., 2000, 2003).

A direct test for lateralisation of language activation using fMRI: comparison with the Wada test

Measurement of language activation using fMRl has become an important tool for assessment of language lateralization in patients undergoing neurosurgerical evaluation. Most studies have computed a lateral&y index (LI) based on the number of activated pixel in each hemisphere. However, LI is potentially susceptible to the statistical threshold (p) chosen, with a higher threshold resulting in more lateralized LI values and lower thresholds in more bilateral LI’s. The aim of the present study was to demonstrate that this ambiguity can be reduced by directly comparing the magnitude of task induced activation between homologous voxels in each hemisphere. We applied this procedure to pediatric patients (aged 12-16 years) who underwent assessment of hemispheric dominance for language using the intracarotid sodium amytal test (Wada Test).