Introduction
As noted in previous chapters, ASD is a highly heterogeneous neurodevelopmental condition with multiple causes and courses, a wide range in symptom severity and many associated comorbid disorders (e.g. see Amaral and colleagues230) – including ADHD. It has thus been suggested that ASD should be thought of as ‘the autisms’ rather than a single autistic phenotype.231,232 This makes the diagnosis of ASD (and its distinction from ADHD) complex. Moreover, it means that clinical trials (in both disorders) necessarily include heterogeneous samples of individuals who may not share the same biological (or cognitive) deficits. This may account for the failure of many clinical trials in ASD and ADHD. Hence, we need to be able to ‘fractionate’ clinical populations better in order to improve diagnostic accuracy and identify more homogeneous groups for clinical trials. One potential way forward is to use recent advances (gained from neuroimaging) in our understanding of the biology of both disorders.
For instance, there is increasing evidence that both ASD and ADHD are associated with abnormalities in specific brain regions and neural systems.233–236 However, reports of region-specific differences in ASD (and ADHD) are highly variable (reviewed in Toal and colleagues237 and Amaral and colleagues230). Such variable findings may simply be explained by confounds such as clinical heterogeneity between studies or analytical techniques. Alternatively, variability in findings may indicate that differences in brain anatomy in ASD and ADHD are relatively subtle and spatially distributed, and are difficult to detect using mass-univariate (i.e. voxel-wise) approaches. Finally, given the likely multiaetiology of ASD and ADHD, it is likely that their neuroanatomy is not confined to a single morphological parameter but affects multiple cortical features.
There is already evidence to suggest that several aspects of cerebral morphology are different in people with ASD and ADHD – including both volumetric (i.e. cortical thickness, regional area) and geometric features (i.e. cortical shape)238,239 – and that different morphological features may have different neuropathological and genetic underpinnings.240 So far, individual aspects of the cortex are generally explored in isolation (i.e. in separate statistical models). Hence, it is not yet known (1) if they equally contribute to differentiating individuals with ASD and ADHD from controls and (2) what the relationship between such multiple cortical abnormalities in ASD is. In the present study, we therefore attempted to establish proof of concept that a multiparameter classification approach employing a support vector machine (SVM) (see Mourão-Miranda and colleagues,241 Davatzikos and colleagues242 and Kloppel and colleagues 243), to investigate brain anatomy in adults with ASD and ADHD. Here, a variety of morphological features, including both volumetric and geometric parameters, were used simultaneously to classify control and ASD/ADHD individuals.
We aimed to demonstrate that the neuroanatomical patterns discriminating individuals with autism and ADHD from controls are truly multidimensional, comprising multiple and most likely independent cortical features. If so, this may guide further exploration of the specific genetic and neuropathological underpinnings of both disorders and provide proof of concept that these may be used to help classify individuals with and without the disorders.
Materials and methods
Participants
Twenty control adults were recruited locally by advertisement and 20 adults with ASD were recruited through a clinical research programme at the Maudsley Hospital/IoP, London. All volunteers () gave informed consent (as approved by the IoP and Bethlem and Maudsley Hospital Trust REC), and had a full-scale IQ of > 75 (WASI244).
All volunteers were right-handed males and were between 20 and 68 years of age. None had a history of major psychiatric disorder or medical illness affecting brain function (e.g. psychosis or epilepsy). All participants underwent a psychiatric interview and physical examination. Blood tests were used to exclude biochemical, haematological or chromosomal abnormalities (including fragile X syndrome). All participants with ASD were diagnosed with autism according to ICD-10 research criteria245 by a trained and qualified clinician in our adult autism specialist clinic at the South London and Maudsley Hospital. The initial clinical diagnosis was then confirmed using the ADI-R,27 whenever possible (i.e. where parents agreed to take part, n = 17). In 3 out of the 20 included cases, ADI-R scores could not be obtained because informants were unavailable. In these three cases, the clinical diagnosis was confirmed using the ADOS.196 Both ADI-R and ADOS-G scores were available for two participants. All cases had to reach the ADOS-G or ADI-R algorithm cut-off points in the three domains of impaired reciprocal social interaction, communication and repetitive behaviours and stereotyped patterns, but we did allow failure to reach cut-off point in one of the domains (by one point). Sixteen ASD individuals did not have a delay in development of phrase speech at the age of 36 months (and so may be subtyped by some as having the autistic subtype of Asperger syndrome); the remaining four individuals had a history of delayed phrase speech (and so may be subtyped by some as having the autistic subtype of high-functioning autism). However, due to the small sample size it was not possible to reliably investigate putative differences between people with HFA and Asperger syndrome, and all subjects were analysed in a combined group of individuals with ASD.
Furthermore, we recruited a group of 19 individuals with ADHD from the adult ADHD services at the Maudsley Hospital, London. This group served as a neurodevelopmental control group and was matched to the ASD group in sex, age (mean 31 ± 8.5 years), full-scale IQ (mean 1.7 ± 6) and handedness. Of these, eight individuals fulfilled the criteria for ADHD combined type and 11 for ADHD inattentive type. The diagnosis of ADHD was made by a trained and qualified clinician based on a structured clinical interview according to DSM-IV criteria. ADHD symptoms were also assessed using Barkley Current and Childhood Symptoms Scales207 and the Wender Utah Rating Scale.246
Magnetic resonance imaging data acquisition
Magnetic resonance imaging (MRI) data were acquired using a 1.5-T GE Signa Neuro-optimized System (General Electric Medical Systems, Milwaukee, WI, USA) fitted with 40 mT/m high-speed gradients at the Maudsley Hospital, London. Whole-brain spoiled gradient recalled acquisition in the steady-state T1-weighted series were collected in the coronal plane with repetition time = 13.8 months, echo time = 2.8 months, yielding 124 contiguous 1.5-mm2 axial slices of 256 × 192 voxels with an in-plane resolution of 1 mm2. Similar T1-weighted MRI scans were acquired for the ADHD group on a 1.5-T GE Signa Neuro-optimized System at the Maudsley Hospital, London, using repetition time = 10.72 months, echo time = 4.86 months, yielding 146 contiguous 1.1-mm2 axial slices. A quadrature birdcage head coil was used for radiofrequency transmission and reception. Foam padding and a forehead strap were used to limit head motion. All scans were visually inspected by a radiologist to assess (1) image contrast, (2) movement artefacts and (3) the existence of clinical abnormalities. Scans displaying low image quality or clinical abnormalities were excluded from this study.
Image processing
Cortical reconstruction and volumetric segmentation was performed with the FreeSurfer image analysis suite (Laboratory for Computational Neuroimaging, Charlestown, MA, USA), which is documented and freely available for download online [URL: http://surfer.nmr.mgh.harvard.edu/ (accessed 31 May 2018)]. The technical details of these procedures are described in prior publications.247–254 Briefly, this processing includes motion correction and averaging of multiple volumetric T1-weighted images (when more than one is available), removal of non-brain tissue using a hybrid watershed/surface deformation procedure,249 automated Talairach transformation, segmentation of the subcortical white matter and deep grey matter volumetric structures (including hippocampus, amygdala, caudate, putamen, ventricles),251,252 intensity normalisation, tessellation of the grey matter–white matter boundary, automated topology correction,255,256 and surface deformation following intensity gradients to optimally place the grey/white and grey/cerebrospinal fluid borders at the location where the greatest shift in intensity defines the transition to the other tissue class.247 Once the cortical models are complete, a number of deformable procedures can be performed for further data processing and analysis, including surface inflation,247 registration to a spherical atlas (which utilised individual cortical folding patterns to match cortical geometry across subjects248), parcellation of the cerebral cortex into units based on gyral and sulcal structure253,257 and creation of a variety of surface-based data, including maps of curvature and sulcal depth. This method uses both intensity and continuity information from the entire three-dimensional magnetic resonance volume in segmentation and deformation procedures to produce representations of cortical thickness, calculated as the closest distance from the grey/white boundary to the grey/cerebral spinal fluid boundary at each vertex on the tessellated surface.250 The maps are created using spatial intensity gradients across tissue classes and are therefore not simply reliant on absolute signal intensity. The maps produced are not restricted to the voxel resolution of the original data, thus are capable of detecting submillimetre differences between groups. Procedures for the measurement of cortical thickness have been validated against histological analysis258 and manual measurements.259,260 FreeSurfer morphometric procedures have been demonstrated to show good test–retest reliability across scanner manufacturers and across field strengths.261 All reconstructed surfaces were visually inspected for gross anatomical topological defects. Brains that did not reconstruct or showed geometric inaccuracies were excluded from the study.
A set of five morphometric parameters per vertex were used as input to the multimodal classifier. Three parameters [1–3: average convexity or concavity, mean (radial) curvature and metric distortion] accounted for geometric features at each cerebral vertex and two parameters (4–5: cortical thickness and surface area, respectively), measured volumetric features.
The average convexity or concavity was used for quantifying the primary folding pattern of a surface. Convexity/concavity captures large-scale geometric features and is insensitive to noise in the form of small wrinkles in a surface.248 At each vertex, convexity/concavity indicates the depth/height above the average surface and measures sulcal depth or gyral height, respectively. Differences in sulcal depth have previously been investigated as one important aspect of the cerebral geometry implicated in ASD238 and have been suggested to reflect abnormal pattern of cortical connectivity. Mean (radial) curvature was used to assess folding of the small secondary and tertiary folds in the surface. The selection of this feature was motivated by early findings of polymicrogyria (i.e. excessive number of small convolutions on the surface of the brain) in autism.262 These would not be captured by the average convexity measure reflecting large-scale geometric features only. Metric distortion (i.e. Jacobian) indicating the degree of cortical folding was calculated as degree of displacement and convolution of the cortical surface relative to the average template.248,263 Although sulcal depth and radial curvature measure specific aspects of the cortical geometry, metric distortion is a wider measure of the overall degree of cortical folding and thus captures geometric distortions otherwise not specified. There is a large body of evidence to suggest that individuals with ASD display abnormal patterns of cortical gyrification reflecting cerebral development and connectivity (e.g. Levitt and colleagues239 and Hardan and colleagues264). Cortical thickness (see above) and pial area (i.e. the area of a vertex on the grey matter surface, calculated as the average of the area of the triangles touching that vertex) were used to quantify volumetric differences. Although these two features are generally combined to measure regional brain volume, a recent study has shown that cortical thickness and surface are influenced by distinct genetic mechanism240 and we therefore added as separate features into the model. A summary of the set of parameters is displayed in .
Summary of the five morphometric parameters measured at each cerebral vertex. (a), Average convexity; (b) cortical thickness; (c) pial area; (d) metric distortion (Jacobian); and (e) mean (radial) curvature. g/w, grey/white. Adapted from Ecker et al. (more...)
These included average convexity, cortical thickness, pial area, metric distortion (Jacobian) and mean (radial) curvature.
Group differences in intracranial volume, total brain volume and grey matter volume as estimated by FreeSurfer were assessed using t-tests for independent samples prior to classification. There were no significant differences between groups in any of these parameters at a level of p < 0.05 (). There were also no significant differences in total brain volume between controls and individuals with ADHD, nor between ASD and ADHD group on p < 0.05. Bartlett’s test for homogeneity of variances was used to examine parameter variability across homologue regions in different hemispheres using a threshold of p < 0.0014 (corrected for multiple comparisons).
Between-group differences in overall brain volume and grey matter volume
Classification and support vector machine
A linear SVM was used to classify between patient and control group on the basis of their brain morphology. SVM has previously been applied to MRI data241–243 and a detailed description can be found in Schoelkopf and Smola265 and Burges.266 Briefly, SVM is a supervised multivariate classification method that treats each image as a point in a high-dimensional space. If SVM is applied to images coming from different modalities, as in the current study, the number of dimensions equals the number of voxels/vertices per image multiplied by the number of modalities. Input data were then classified into two classes (e.g. individuals with ASD and controls), by identifying a separating hyperplane or decision boundary. The algorithm is initially trained on a subset of the data <x,c> to find a hyperplane that best separates the input space according to the class labels c (e.g. –1 for patients, +1 for controls). Here, x represents the input data (i.e. feature vector). The feature vector was generated by concatenating the five image modalities for each subject. Once the decision function is learned from the training data, it can be used to predict the class of a new test example. SVM267 is a maximum margin classifier, which identifies the optimal hyperplane by finding the hyperplane with the maximum margin (i.e. maximal separation between classes). The margin is the distance from the separating hyperplane to the closest training examples. The training examples that lie on the margin are called support vectors. The hyperplane is defined by a weight vector and an offset. The weight vector is a linear combination of the support vectors and is normal to the hyperplane. The linear kernel SVM used in the present study allows direct extraction of the weight vector as an image (i.e. the SVM discrimination map). A parameter C, that controls the trade-off between having zero training errors and allowing misclassifications, was fixed at C = 1 for all cases (default value). The LIBSVM toolbox for MATLAB (ACM Transactions on Intelligent Systems and Technology, Taiwan) was used to perform the classifications [URL: www.csie.ntu.edu.tw/∼cjlin/libsvm/ (accessed 31 May 2018)].
Discrimination maps
The weight vector has the same dimension as the feature vector and is normal to the hyperplane. It can be thought of as a spatial representation of the decision boundary and thus represents a map of the most discriminating regions. Here, the feature vector had n × m dimensions, where n is the number of vertices and m denotes the number of modalities (i.e. five). Given two groups (ASD vs. controls), with the labels +1 and –1, respectively, a positive value in the discrimination map (red/yellow colour scale) indicates relatively higher parameter values in patients than in controls with respect to the hyperplane, and a negative weight (blue/cyan colour scale) means relatively higher parameter values in controls than in patients with respect to the hyperplane. As the classifier is multivariate by nature, the combination of all voxels as a whole is identified as a global spatial pattern by which the groups differ (i.e. the discriminating pattern). To enable visualisation of the discriminating pattern for each image modality, the weight vector was cut into its constituent parts, which were then mapped back onto the average white matter surface. In the present study we coloured all voxels that have values > 30% of the maximum value of the discrimination map.241,268 This threshold, although ultimately arbitrary, eliminates noise components predominantly (< 30%), thus enabling a better visualisation of the most discriminating regions (31–100%).
Intraregional morphometric profiles
To identify the relative contribution of specific parameters in discriminating between groups at different locations on the cortical surface, we displayed so called intraregional morphometric profiles. These profiles were derived by calculating the average weights of vertices within regions of interest (ROI) for the five different parameters. ROIs were based on contiguous weight clusters from the overall discrimination maps. The choice of the ROIs, although ultimately arbitrary, was motivated by (1) the relevance of a region to the current literature (e.g. sulcal depth differences in intraparietal sulcus238) and (2) generally high parameter weights for a specific morphometric feature (e.g. high weights for cortical thickness in medial temporal sulcus). The ROI analysis aimed to illustrate that different regions can display distinct differences in one or more parameters, rather than displaying the same pattern of feature weights. As all weights were scaled, means and SDs are directly comparable across regions and parameters.
Cross-validation and permutation testing
The performance of the classifier was validated using the commonly used leave-two-out cross-validation approach. This validation approach provides robust parameter estimates particularly for smaller samples. In each trial observations from all but one subject from each group were used to train the classifier. Subsequently, the class assignment of the test subjects was calculated during the test phase. This procedure was repeated S = 20 times (where S is the number of subjects per group), each time leaving observations from a different subject from each group out. The accuracy of the classifier was measured by the proportion of observations that were correctly classified into patient or control group. We also quantified the sensitivity and specificity of the classifier defined as:
where TP is the number of true positives (i.e. the number of patient images correctly classified); TN is the number of true negatives (i.e. number of control images correctly classified); FP is the number of false positives (i.e. number of controls images classified as patients); and FN is the number of false negatives (i.e. number of patients images classified as controls).
Classifications were made, first, by including all five morphological modalities into the feature vector and, second, on the basis of each individual parameter. Different classifiers were trained for each hemisphere. This enabled us to assess the overall classification accuracies for individual parameters and hemispheres. Classifier performance was evaluated using basic ROC graphs as well as permutation testing. Permutation testing can be used to evaluate the probability of getting specificity and sensitivity values higher than the ones obtained during the cross-validation procedure by chance. We permuted the labels 1000 times without replacement, each time randomly assigning patient and control labels to each image and repeated the cross-validation procedure. We then counted the number of times the specificity and sensitivity for the permuted labels were higher than the ones obtained for the real labels. Dividing this number by 1000 we derived a p-value for the classification.
Finally, the established classifier was used to predict group membership of individuals with ADHD. To achieve this, the classifier was first retrained using all ASD–control pairs to provide the best possible prediction model for discriminating the two groups. This trained classifier was then applied to the multiparameter data from the ADHD group and predictions for group membership for these subjects were derived.
To validate the binary classification of the subject groups on a quantitative level and to identify the degree to which the classification is driven by autistic symptoms rather than confounds unrelated to autism, the test margin for each subject coming from the all included classifier was correlated with the level of symptom severity measured by the ADI-R subscales. A similar approach has previously been employed by Ecker et al.269
Results
Overall classifier performance
Classification accuracies as well as sensitivity and specificity for each classifier are listed in . On the whole, strong hemispheric asymmetry was observed with regard to the overall classification accuracy. The left hemisphere provided consistently higher and above chance prediction accuracies across all morphometric parameters. Here, individuals with ASD were correctly assigned to the appropriate diagnostic category in 85.0% of all cases when all parameters were considered simultaneously (). The sensitivity of the multiparameter classification in the left hemisphere was 90.0% (i.e. if a volunteer had a clinical diagnosis of ASD, the probability that this participant was correctly assigned to the ASD category was 0.9). The specificity was 80.0%, meaning that 80.0% of the controls subjects were correctly classified as controls. High classification accuracies were also obtained when individual morphological parameters were considered. shows the receiver operator characteristics or ROC graph for the six classifiers in the left hemisphere. In general, one point in ROC space is better than another if it is to the north-west (true-positive rate is high, false-positive rate is low, or both), with the point (0,1) representing perfect classification. Classifiers appearing on the left-hand side of the ROC graph, near the x-axis, may be considered ‘conservative’ (i.e. make positive classifications only with strong evidence but often have low true-positive rates), whereas classifiers on the upper right-hand side of the graph may be thought of as ‘liberal’ (i.e. make positive classifications with weak evidence so they classify nearly all positives correctly but often have high false positives). The diagonal line y = x represents the strategy of randomly guessing a class. Best discrimination was obtained when cortical thickness measures were used to classify between groups with sensitivity and specificity values as high as 90.0%, followed by a classification on the basis of the metric distortion parameter (sensitivity 80.0%, specificity 80.0%). The average convexity, pial area and mean curvature displayed accuracies in the range of 70.0–80.0%. The classification p-value resulting from the permutation test was very low across, as well as within, modalities (< 0.004). The probability of obtaining specificity and sensitivity values higher than the ones obtained during cross-validation procedure by chance is thus extremely low.
Results of SVM classification between ASD and control group using different brain morphometric features in the left and right hemisphere
Receiver operating characteristic graphs for classifiers in the left hemisphere and right hemisphere. (a) Left hemisphere; (b) right hemisphere; (c) classification plots for the left hemisphere; and (d) classification plots for the right hemisphere. A, (more...)
A very similar profile of parameter importance was observed in the right hemisphere – despite it having generally lower classification accuracies. Here, individuals with ASD were correctly assigned to the appropriate diagnostic category in 65.0% of all cases (sensitivity 60.0%, specificity 70.0%; p < 0.03), when all parameters were considered simultaneously (see and ). As in the left hemisphere, cortical thickness as well as metric distortion displayed the best classification performance, although only cortical thickness measures reached statistical significance (p < 0.01). The discrepancy in overall classification accuracy between the two hemispheres was not due to differences in parameter variability, as there were no significant differences in parameter variance between homologous regions in different hemispheres (calculated across the whole sample).
Thus, although all parameters provided statistically significant predictions in the left hemisphere, only the full model and cortical thickness displayed significant predictions in the right hemisphere. Highest classification accuracy (90.0%) might thus be obtained by using cortical thickness measures exclusively in the left hemispheres. Correlation coefficients between ADI-R subscales and the test margin coming from the combined model are listed in . The test margin in the left hemisphere was positively correlated with the ADI-R scores for individuals with ASD in the social (r = 0.414; p < 0.05) and communication domain (r = 0.62; p < 0.01). Therefore, individuals with higher values on these ADI-R subdomains are located on the extreme right relative on the hyperplane, whereas the individuals with a lower level of impairment are mostly located in close proximity to the hyperplane overall.
Correlation coefficients between ADI-R diagnostic criteria and weight vector for the model combining all parameters
Classification of individuals with ADHD using the established ASD classifier
In order to establish the degree of clinical specificity of the discrimination algorithm to ASD, rather than to NDs in general, the established classifier was used to predict group membership of individuals with ADHD. To predict group membership, the model including all parameters was chosen as it provided best classification accuracy in the right hemisphere and high accuracy in the left hemisphere. On the basis of the neuroanatomical information available for the left hemisphere, 15 out of the 19 individuals with ADHD were allocated to the control group (78.9%) and four individuals with ADHD were allocated to the ASD group (21%). Using the right hemisphere, the classifier allocated cases with ADHD with approximately equal frequencies (47% allocated to the ASD category; 52% allocated to the control category). The classification plots for individuals with ADHD are shown in for both hemispheres.
Classification plots showing group allocation of individuals with ADHD (blue squares) in the (a) left and (b) right hemisphere using the ASD classifier. Reproduced from Ecker et al. © 2010 Ecker et al.; licensee BioMed Central Ltd. This article (more...)
Discrimination maps of ASD-specific abnormalities
The spatial maps of brain regions identified as outlined above are shown in and a summary description can be found in . Each of these maps is a spatial representation of the SVM weight vector, and although they do not directly quantify the information content of each region, each map forms a spatially distributed pattern showing the relative contribution of each voxel to the decision function. Although ‘knockout’ techniques may be used to directly quantify the information content of particular regions, this approach may be hampered by substantial information redundancy across brain regions. Nonetheless, the spatial distribution of the weight vector does provide information about which brain regions contributed to classification and in this case is suggestive of a distributed pattern of relative deficit or excess in ASD with respect to controls. We emphasise that due to the multivariate character of SVM (i.e. it considers inter-regional correlations), each region in the discrimination maps should be interpreted in the context of the entire discriminating pattern and should not be considered in isolation. Furthermore, individual regions may display high classification weights for several reasons [e.g. there is a large difference in volume between groups in that region, or the region is highly intercorrelated with other components of the network (i.e. pattern)]. This is of particular relevance to ASD, as individuals with the disorder most likely have abnormalities in the development of neural systems, in addition to differences in isolated regions. Thus, it is important to highlight that individual network components are not necessarily different between groups, but should be considered as constituent parts of a neuroanatomical network discriminating between groups.
Discrimination maps for the five different morphometric features in the left and right hemispheres. Colour maps represent the weight vector on the basis of the five modality classification for (a) cortical thickness; (b) average convexity; (c) metric (more...)
Regions displaying high discrimination weights between ASD and control group for volumetric measures
As can be seen in , the five investigated parameters resulted in different spatially distributed patterns of regions with highest contribution to the discrimination. Although cortical thickness provided best classification accuracy and maximum weight values, differences between groups were observed in both volumetric as well as geometric features.
Colour maps represent the weight vector on the basis of the five modality classification for cortical thickness (see ), average convexity (see ), metric distortion (see ) and pial area (see ). Weights for the mean (radial) curvature did not exceed the set threshold. Positive weights (i.e. overall excess patters in ASD relative to controls) are displayed in red, negative weights (i.e. overall deficit patterns in ASD relative to controls) are displayed in blue (see ).
Cortical thickness
In both hemispheres, the discriminative pattern for cortical thickness in ASD comprised regions in all four lobes of the cortex (see ). Regional details are summarised in . The ‘excess pattern’ (i.e. ASD > controls) comprised several temporal regions, including the right superior temporal sulcus (STS, BA22), medial and superior temporal gyrus (BA21/BA22) as well as the parahippocampal gyrus (BA36), fusiform gyrus (BA20) and enthorinal cortex. In addition, the excess pattern included the inferior/superior parietal lobe (BA39), Brodmann area 18 of the occipital lobe, and the anterior and posterior cingulate gyrus.
Although the excess pattern comprised predominantly occipito-temporal regions, the pattern displaying a relative thinning of the cortex in ASD versus controls (i.e. ‘deficit pattern’) included mainly frontal and parietal regions such as the middle frontal gyrus (BA46), the medial/superior frontal gyrus (BA10) and BA46 in the medial frontal gyrus. In addition, the deficit pattern contained the superior parietal cortex (BA40/7) and in the anterior cingulate gyrus.
Surface area
Only a few surface area (see ) differences were observed in relatively small clusters with generally low weights. Details are summarised in .
Regional geometric characteristics
Apart from volumetric differences, features describing regional geometric characteristics also displayed high discrimination weights, which were summarised in . In both hemispheres, individuals with ASD displayed a pattern of relative increase in sulcal depth (see ) in the intraparietal sulcus as well as in the superior frontal cortex. The discriminative pattern for cortical folding (see ), as indicated by the metric distortion (i.e. Jacobian), included mainly bilateral parietal regions such as the inferior parietal lobe (BA39/40), and several regions of the right frontal lobe (e.g. supramarginal gyrus, postcentral gyrus, orbitofrontal regions). In addition, the discriminative pattern in cortical folding comprised the precuneus.
Regions displaying high discriminative weights between ASD and control group for geometric measures
Morphometric profiles
The discrimination maps show that different morphometric parameters elicited different spatial patterns of weights with some overlap between particular parameter pairs. The weights for individual parameters are thus not uniform across the cortex, but vary from region to region (i.e. vertex to vertex). This can also be seen on the basis of the morphometric profile for individual regions, which are displayed in – and listed in . These profiles visualise the mean parameter weights for individual features within ROI.
Visualisation of the morphometric abnormalities in the right intraparietal sulcus. (a) Colour maps represent the weight vector score; (b) outlines of the cortical surface for ASD (red) and control (blue) group (this main discriminating factor in this (more...)
(a) Morphometric abnormalities in the posterior cingulate gyrus; (b) and (c) a combination of cortical thickness and folding pattern lead to a high contribution to the classification in that region. Reproduced from Ecker et al. © 2010 Ecker et (more...)
Mean discrimination weights within ROI for individual morphometric features
(a) Morphometric abnormalities in the middle temporal sulcus; (b) visualisation of cortical thickness measures within the middle temporal sulcus; (c) and (d) in this region, only cortical thickness discriminated between ASD and control groups, for which (more...)
In the intraparietal sulcus for instance, high discrimination weights were observed for differences in sulcal depth with lower weights for volumetric features such as cortical thickness or surface area (see ). A similar profile was also seen in the inferior parietal lobe (BA39). Here, the weights for volumetric parameters such as cortical thickness or surface area were very low in comparison with the weights associated with the pattern of cortical folding (i.e. metric distortion) (see ).
(a) Visualisation of the morphometric abnormalities in the left inferior parietal lobe (BA39); (b) outlines of the cortical surface for ASD and control group; (c) differences in metric distortion for this ROI are shown for both groups; and (d) morphometric (more...)
Other regions, however, displayed high weights for volumetric parameters exclusively. For instance, in the temporal sulcus (BA21), high weights were observed for cortical thickness exclusively in the light of low weights in all other parameters (see ). There were also regions with equally high weights in volumetric as well as geometric features (e.g. the anterior cingulate gyrus). Here, cortical thickness as well as local cortical folding displayed high parameter weights.
In summary, we observed a spatially distributed pattern of regions that can be used to differentiate individuals with ASD from controls. Overall, highest weights were observed for measures of cortical thickness, followed by geometric features such as sulcal depth and metric distortion, which were predominantly observed in parietal regions. This suggests that both volumetric and geometric features play an important role in distinguishing between ASD and control group. Finally, we demonstrated that the parameter weights for individual cortical features were region dependent.
Discussion
Here, we used a multiparameter classification approach to characterise the complex and subtle grey matter differences in adults with ASD and ADHD. SVM achieved good separation between groups and revealed spatially distributed and largely non-overlapping patterns of regions with highest classification weights for each of five morphological features. Our results confirm that the neuroanatomy of ASD and ADHD are truly multidimensional, affecting multiple neural systems. The discriminating patterns detected using SVM may help classify affected individuals.
There is good evidence to suggest that several aspects of cerebral morphology are implemented in ASD and ADHD, including both volumetric and geometric features.238,239 However, these are normally explored in isolation. Here, we aimed to establish a framework for multiparameter image classification to describe differences in grey matter neuroanatomy in ASD and ADHD in multiple dimensions, and to explore the predictive power of individual parameters for group membership. This was achieved using a multiparameter classifier incorporating volumetric and geometric features at each cerebral vertex. In the left hemisphere, SVM correctly classified 85% of all cases overall at a sensitivity and specificity as high as 90% and 80%, respectively, using all five morphological features. This level of sensitivity compares well with behaviourally guided diagnostic tools, whose accuracies are on average around 80%. Naturally, one would expect lower sensitivity values than the test used for defining the ‘autistic prototype’ itself (i.e. ADI-R). Thus, if a classifier is trained on the basis of true positives identified by diagnostic tools, the maximal classification accuracy that could be reached is only as good as the measurements used to identify true positives.
The significant predictive value of pattern classification approaches may have potential clinical applications. Currently, both ASD and ADHD are diagnosed solely on the basis of behavioural criteria. The behavioural diagnosis is, however, often time-consuming and can be problematic, particularly in adults. In addition, different biological aetiologies might result in the same behavioural phenotype (the ‘autisms’232), which is undetectable using behavioural measures alone. Thus, the existence of an ASD (or ADHD) biomarker such as brain anatomy might be useful to facilitate and guide the behavioural diagnosis. This would, however, require further extensive exploration in the clinical setting, particularly with regard to classifier specificity to ASD or ADHD – rather than neurodevelopmental conditions in general.
To address the issue of clinical specificity, the established ASD classifier was used to classify individuals with ADHD. Bilaterally, the ASD classifier did not allocate the majority of ADHD subjects to the ASD category. This indicates that it does not perform equally well for other neurodevelopmental conditions and is more specific to ASD. To further demonstrate that the classification is driven by autistic symptoms, the test margins of individuals with ASD were correlated with measures of symptom severity.269 We found that larger margins were associated with more severe impairments in the social and communication domain of the ADI-R. The classifier therefore seems to utilise neuroanatomical information specifically related to ASD rather than simply reflecting non-specific effects introduced by any kind of pathology. However, due to a recent scanner upgrade, ADHD scans were acquired with different acquisition parameters, while manufacturer, field strength and pulse sequence remained the same. FreeSurfer has been demonstrated to show good test–retest reliability particularly within scanner manufacturer and field strength,261 but we cannot exclude the possibility that systematic differences in regional contrast may have affected the ADHD classification. Future research is thus needed to validate the ADHD findings on an independent sample.
The overall classification accuracy varied across hemispheres (79.0% left vs. 65.0% right) in the absence of interhemispheric differences in parameter variability. Hemisphere laterality is an area, which remains relatively unexplored in autism. Although our data suggest that the left hemisphere is better at discriminating between groups (i.e. is more ‘abnormal’), it is unclear whether this discrepancy is due to quantitative differences in parameters or to qualitative aspects of the discriminating patterns (i.e. additional regions). Furthermore, it is also not possible to identify whether or not individuals with ASD or ADHD display a higher (lower) degree of cortical asymmetry relative to controls. There is some evidence to suggest that individuals with ASD show a lower degree of ‘leftward’ (i.e. left > right) cortical symmetry than controls,270 which may explain differences in classification accuracy. There is also evidence to suggest that the left hemisphere is under tighter genetic control than the right hemisphere,271 which may be of relevance to a highly heritable condition such as ASD. However, a direct numerical comparison between hemispheres is needed to address this issue directly.
The classification accuracy varied not only across hemispheres, but also across morphometric parameters. Bilaterally, cortical thickness provided the best classification accuracy and highest regional weights. Differences in cortical thickness have been reported previously in ASD and ADHD for both increases272,273 as well as decreases,272,274 and in similar regions as reported here (i.e. parietal, temporal and frontal areas). The overlap with previous studies indicates that these regions display high classification weights due to a quantitative (i.e. ‘true’) difference rather than high intercorrelations with thickness measures in other brain regions.
Certain geometric features such as average convexity and metric distortion provided above chance classifications as well, particularly in parietal, temporal and frontal regions, and in areas of the cingulum. Average convexity and metric distortion measure different aspects of cortical geometry (see Materials and methods) and have previously been linked to ASD, as has sulcal depth.238 Such geometric features were suggested to reflect abnormal patterns of cortical connectivity. There have also been reports of abnormal patterns of gyrification262,264 and large-scale displacements of the major sulci.239 Thus, our study provides further evidence to support the hypothesis that the ‘autistic brain’ is not just bigger or smaller but is also abnormally shaped and that this is ‘separable’ from ADHD.
Strengths and weaknesses
Our findings should be interpreted in the context of a number of methodological limitations.
First, the classification algorithm is highly specific to the particular sample used for ‘training’ the classifier, namely high-functioning adults with ASD. The advantage of this approach is that the classifier offers high specificity with regard to this particular subject group, but is less specific to other cohorts on the spectrum. Owing to the small sample size it was also not possible to reliably investigate differences between HFA and Asperger syndrome. Evidence275 suggests that by adulthood these groups are largely indistinguishable at the phenotypic level. However, the extent to which these groups differ at the level of brain anatomy is unknown and may be worth investigating using SVM in the future. Second, 85% of ASD participants in our sample were diagnosed using the ADI-R and 15% were diagnosed using the ADOS-G. As both diagnostic tools measure autistic symptoms at different developmental stages, the classifier may be biased towards individuals with an early diagnosis of ASD. Although it is not expected that classifier performance on the basis of ADOS-G and ADI-R differ drastically, diagnostic heterogeneity may be a potential limitation. Finally, SVM is a multivariate technique and hence offers a limited degree of interpretability of specific network components. Additional analysis such as ‘searchlight’ or ‘virtual lesions’ approaches276–278 may therefore be combined with SVM in the future to establish the relative contribution of individual regions/parameters to the overall classification performance.
