Combining multivariate voxel selection and support vector machines for mapping and classification of fMRI spatial patterns

A Brainnetome Atlas Based Mild Cognitive Impairment Identification Using Hurst Exponent

Mild cognitive impairment (MCI), which generally represents the transition state between normal aging and the early changes related to Alzheimer’s disease (AD), has drawn increasing attention from neuroscientists due that efficient AD treatments need early initiation ahead of irreversible brain tissue damage. Thus effective MCI identification methods are desperately needed, which may be of great importance for the clinical intervention of AD. In this article, the range scaled analysis, which could effectively detect the temporal complexity of a time series, was utilized to calculate the Hurst exponent (HE) of functional magnetic resonance imaging (fMRI) data at a voxel level from 64 MCI patients and 60 healthy controls (HCs). Then the average HE values of each region of interest (ROI) in brainnetome atlas were extracted and compared between MCI and HC. At last, the abnormal average HE values were adopted as the classification features for a proposed support vector machine (SVM) based identification algorithm, and the classification performance was estimated with leave-one-out cross-validation (LOOCV). Our results indicated 83.1% accuracy, 82.8% sensitivity and 83.3% specificity, and an area under curve of 0.88, suggesting that the HE index could serve as an effective feature for the MCI identification. Furthermore, the abnormal HE brain regions in MCI were predominately involved in left middle frontal gyrus, right hippocampus, bilateral parahippocampal gyrus, bilateral amygdala, left cingulate gyrus, left insular gyrus, left fusiform gyrus, left superior parietal gyrus, left orbital gyrus and left basal ganglia.

The multidimensional representational space of observed socio-affective touch experiences

Observed touch interactions provide useful information on how others communicate with the external world. Previous studies revealed shared neural circuits between the direct experience and the passive observation of simple touch, such as being stroked/slapped. Here, we investigate the complexity of the neural representations underlying the understanding of others' socio-affective touch interactions. Importantly, we use a recently developed touch database that contains a larger range of more complex social and non-social touch interactions. Participants judged affective aspects of each touch event and were scanned while watching the same videos. Using correlational multivariate pattern analysis methods, we obtained neural similarity matrices in 18 regions of interest from five different networks: somatosensory, pain, the theory of mind, visual and motor regions. Among them, four networks except motor cortex represent the social nature of the touch, whereas fine-detailed affective information is reflected in more targeted areas such as social brain regions and somatosensory cortex. Lastly, individual social touch preference at the behavioral level was correlated with the involvement of somatosensory areas on representing affective information, suggesting that individuals with higher social touch preference exhibit stronger vicarious emotional responses to others' social touch experiences. Together, these results highlight the overall complexity and the individual modulation of the distributed neural representations underlying the processing of observed socio-affective touch.

•

∙Neural bases of observed touch are investigated with touch videos and MVPA.

•

∙Social touch evokes stronger activation in the theory of mind (ToM) network.

•

∙The ToM network represents affective meanings of observed social touch events.

•

∙Affective representations of observed touch are present in somatosensory areas.

•

∙Affective representations in S1 relate to individual's attitude towards touch.

Multi-Site Diagnostic Classification of Schizophrenia Using Discriminant Deep Learning with Functional Connectivity MRI

BACKGROUND

A lack of a sufficiently large sample at single sites causes poor generalizability in automatic diagnosis classification of heterogeneous psychiatric disorders such as schizophrenia based on brain imaging scans. Advanced deep learning methods may be capable of learning subtle hidden patterns from high dimensional imaging data, overcome potential site-related variation, and achieve reproducible cross-site classification. However, deep learning-based cross-site transfer classification, despite less imaging site-specificity and more generalizability of diagnostic models, has not been investigated in schizophrenia.

METHODS

A large multi-site functional MRI sample (n = 734, including 357 schizophrenic patients from seven imaging resources) was collected, and a deep discriminant autoencoder network, aimed at learning imaging site-shared functional connectivity features, was developed to discriminate schizophrenic individuals from healthy controls.

FINDINGS

Accuracies of approximately 85·0% and 81·0% were obtained in multi-site pooling classification and leave-site-out transfer classification, respectively. The learned functional connectivity features revealed dysregulation of the cortical-striatal-cerebellar circuit in schizophrenia, and the most discriminating functional connections were primarily located within and across the default, salience, and control networks.

INTERPRETATION

The findings imply that dysfunctional integration of the cortical-striatal-cerebellar circuit across the default, salience, and control networks may play an important role in the "disconnectivity" model underlying the pathophysiology of schizophrenia. The proposed discriminant deep learning method may be capable of learning reliable connectome patterns and help in understanding the pathophysiology and achieving accurate prediction of schizophrenia across multiple independent imaging sites.

Prediction of activation patterns preceding hallucinations in patients with schizophrenia using machine learning with structured sparsity

Despite significant progress in the field, the detection of fMRI signal changes during hallucinatory events remains difficult and time-consuming. This article first proposes a machine-learning algorithm to automatically identify resting-state fMRI periods that precede hallucinations versus periods that do not. When applied to whole-brain fMRI data, state-of-the-art classification methods, such as support vector machines (SVM), yield dense solutions that are difficult to interpret. We proposed to extend the existing sparse classification methods by taking the spatial structure of brain images into account with structured sparsity using the total variation penalty. Based on this approach, we obtained reliable classifying performances associated with interpretable predictive patterns, composed of two clearly identifiable clusters in speech-related brain regions. The variation in transition-to-hallucination functional patterns not only from one patient to another but also from one occurrence to the next (e.g., also depending on the sensory modalities involved) appeared to be the major difficulty when developing effective classifiers. Consequently, second, this article aimed to characterize the variability within the prehallucination patterns using an extension of principal component analysis with spatial constraints. The principal components (PCs) and the associated basis patterns shed light on the intrinsic structures of the variability present in the dataset. Such results are promising in the scope of innovative fMRI-guided therapy for drug-resistant hallucinations, such as fMRI-based neurofeedback.

Supervised nonlinear dimension reduction of functional magnetic resonance imaging data using Sliced Inverse Regression

Dimension reduction is essential for identifying a small set of discriminative features that are predictive of behavior or cognition from high-dimensional functional magnetic resonance imaging (fMRI) data. However, conventional linear dimension reduction techniques cannot reduce the dimension effectively if the relationship between imaging data and behavioral parameters are nonlinear. In the paper, we proposed a novel supervised dimension reduction technique, named PC-SIR (Principal Component - Sliced Inverse Regression), for analyzing high-dimensional fMRI data. The PC-SIR method is an important extension of the renowned SIR method, which can achieve the effective dimension reduction (e.d.r.) directions even the relationship between class labels and predictors is nonlinear but is unable to handle high-dimensional data. By using PCA prior to SIR to orthogonalize and reduce the predictors, PC-SIR can overcome the limitation of SIR and thus can be used for fMRI data. Simulation showed that PC-SIR can result in a more accurate identification of brain activation as well as better prediction than support vector regression (SVR) and partial least square regression (PLSR). Then, we applied PC-SIR on real fMRI data recorded in a pain stimulation experiment to identify pain-related brain regions and predict the pain perception. Results on 32 subjects showed that PC-SIR can lead to significantly higher prediction accuracy than SVR and PLSR. Therefore, PC-SIR could be a promising dimension reduction technique for multivariate pattern analysis of fMRI.

Human V4 Activity Patterns Predict Behavioral Performance in Imagery of Object Color

Color is special among basic visual features in that it can form a defining part of objects that are engrained in our memory. Whereas most neuroimaging research on human color vision has focused on responses related to external stimulation, the present study investigated how sensory-driven color vision is linked to subjective color perception induced by object imagery. We recorded fMRI activity in male and female volunteers during viewing of abstract color stimuli that were red, green, or yellow in half of the runs. In the other half we asked them to produce mental images of colored, meaningful objects (such as tomato, grapes, banana) corresponding to the same three color categories. Although physically presented color could be decoded from all retinotopically mapped visual areas, only hV4 allowed predicting colors of imagined objects when classifiers were trained on responses to physical colors. Importantly, only neural signal in hV4 was predictive of behavioral performance in the color judgment task on a trial-by-trial basis. The commonality between neural representations of sensory-driven and imagined object color and the behavioral link to neural representations in hV4 identifies area hV4 as a perceptual hub linking externally triggered color vision with color in self-generated object imagery.SIGNIFICANCE STATEMENT Humans experience color not only when visually exploring the outside world, but also in the absence of visual input, for example when remembering, dreaming, and during imagery. It is not known where neural codes for sensory-driven and internally generated hue converge. In the current study we evoked matching subjective color percepts, one driven by physically presented color stimuli, the other by internally generated color imagery. This allowed us to identify area hV4 as the only site where neural codes of corresponding subjective color perception converged regardless of its origin. Color codes in hV4 also predicted behavioral performance in an imagery task, suggesting it forms a perceptual hub for color perception.

Increased Resting-State Cerebellar-Cerebral Functional Connectivity Underlying Chronic Tinnitus

Purpose: Chronic subjective tinnitus may arise from aberrant functional coupling between the cerebellum and the cerebral cortex. To explore this hypothesis, we used resting-state functional magnetic resonance imaging (fMRI) to illuminate the functional connectivity network of the cerebellar regions in chronic tinnitus patients and controls.

Methods: Resting-state fMRI scans were obtained from 28 chronic tinnitus patients and 29 healthy controls (well matched for age, sex and education) in this study. Cerebellar-cerebral functional connectivity was characterized using a seed-based whole-brain correlation method. The resulting cerebellar functional connectivity measures were correlated with each clinical tinnitus characteristic.

Results: Chronic tinnitus patients demonstrated increased functional connectivity between the cerebellum and several cerebral regions, including the superior temporal gyrus (STG), parahippocampal gyrus (PHG), inferior occipital gyrus (IOG), and precentral gyrus. The enhanced functional connectivity between the left cerebellar Lobule VIIb and the right STG was positively correlated with the Tinnitus Handicap Questionnaires (THQ) score (r = 0.577, p = 0.004). Furthermore, the increased functional connectivity between the cerebellar vermis and the right STG was also associated with the THQ score (r = 0.432, p = 0.039).

Conclusions: Chronic tinnitus patients have greater cerebellar functional connectivity to certain cerebral brain regions which is associated with specific tinnitus characteristics. Resting-state cerebellar-cerebral functional connectivity disturbances may play a pivotal role in neuropathological features of tinnitus.

Identifying musical pieces from fMRI data using encoding and decoding models

Encoding models can reveal and decode neural representations in the visual and semantic domains. However, a thorough understanding of how distributed information in auditory cortices and temporal evolution of music contribute to model performance is still lacking in the musical domain. We measured fMRI responses during naturalistic music listening and constructed a two-stage approach that first mapped musical features in auditory cortices and then decoded novel musical pieces. We then probed the influence of stimuli duration (number of time points) and spatial extent (number of voxels) on decoding accuracy. Our approach revealed a linear increase in accuracy with duration and a point of optimal model performance for the spatial extent. We further showed that Shannon entropy is a driving factor, boosting accuracy up to 95% for music with highest information content. These findings provide key insights for future decoding and reconstruction algorithms and open new venues for possible clinical applications.

Perfusion Imaging in Acute Traumatic Brain Injury

Traumatic brain injury (TBI) is a significant problem worldwide and neuroimaging plays a critical role in diagnosis and management. Recently, perfusion neuroimaging techniques have been explored in TBI to determine and characterize potential perfusion neuroimaging biomarkers to aid in diagnosis, treatment, and prognosis. In this article, computed tomography (CT) bolus perfusion, MR imaging bolus perfusion, MR imaging arterial spin labeling perfusion, and xenon CT are reviewed with a focus on their applications in acute TBI. Future research directions are also discussed.

Neural preservation underlies speech improvement from auditory deprivation in young cochlear implant recipients

Although cochlear implantation enables some children to attain age-appropriate speech and language development, communicative delays persist in others, and outcomes are quite variable and difficult to predict, even for children implanted early in life. To understand the neurobiological basis of this variability, we used presurgical neural morphological data obtained from MRI of individual pediatric cochlear implant (CI) candidates implanted younger than 3.5 years to predict variability of their speech-perception improvement after surgery. We first compared neuroanatomical density and spatial pattern similarity of CI candidates to that of age-matched children with normal hearing, which allowed us to detail neuroanatomical networks that were either affected or unaffected by auditory deprivation. This information enables us to build machine-learning models to predict the individual children's speech development following CI. We found that regions of the brain that were unaffected by auditory deprivation, in particular the auditory association and cognitive brain regions, produced the highest accuracy, specificity, and sensitivity in patient classification and the most precise prediction results. These findings suggest that brain areas unaffected by auditory deprivation are critical to developing closer to typical speech outcomes. Moreover, the findings suggest that determination of the type of neural reorganization caused by auditory deprivation before implantation is valuable for predicting post-CI language outcomes for young children.

Feedback from human posterior parietal cortex enables visuospatial category representations as early as primary visual cortex

INTRODUCTION

Categorization is a fundamental cognitive process, whereby the brain assigns meaning to sensory stimuli. Previous studies have found category representations in prefrontal cortex and posterior parietal cortex (PPC). However, these higher-order areas lack the fine-scale spatial representations of early sensory areas, and it remains unclear what mechanisms enable flexible categorization based on fine-scale features.

METHODS

In this study, we decoded functional MRI signals and measured causal influences, across visual, parietal, and prefrontal cortex from participants performing categorization based on coarse- or fine-scale spatial information in thirteen healthy adults.

RESULTS

We show that category information based on coarse discriminations was represented in the PPC, in the intraparietal sulcus region, IPS1/2, at an early stage of categorization trials, whereas representations of category information based on fine-scale discriminations formed later during interactions between IPS1/2 and primary visual cortex (V1). Specifically, when fine-scale discriminations were necessary, we decoded significant category information from V1 at an intermediate stage of trials and again from IPS1/2 at a late stage. IPS1/2 feedback was critical, because categorization performance improved as causal influence from IPS1/2 to V1 increased. Further, these mechanisms were plastic, as the selectivity of IPS1/2 and V1 responses shifted markedly with retraining to categorize the same stimuli into two new groups.

CONCLUSIONS

Our findings suggest that reentrant processing between the PPC and visual cortex enables flexible abstraction of category information.

Identifying ADHD children using hemodynamic responses during a working memory task measured by functional near-infrared spectroscopy

OBJECTIVE

Attention-deficit/hyperactivity disorder (ADHD) is a neurodevelopmental disorder affecting children and adults. Previous studies found that functional near-infrared spectroscopy (fNIRS) can reveal significant group differences in several brain regions between ADHD children and healthy controls during working memory tasks. This study aimed to use fNIRS activation patterns to identify ADHD children from healthy controls.

APPROACH

FNIRS signals from 25 ADHD children and 25 healthy controls performing the n-back task were recorded; then, multivariate pattern analysis was used to discriminate ADHD individuals from healthy controls, and classification performance was evaluated for significance by the permutation test.

MAIN RESULTS

The results showed that 86.0% ([Formula: see text]) of participants can be correctly classified in leave-one-out cross-validation. The most discriminative brain regions included the bilateral dorsolateral prefrontal cortex, inferior medial prefrontal cortex, right posterior prefrontal cortex, and right temporal cortex.

SIGNIFICANCE

This study demonstrated that, in a small sample, multivariate pattern analysis can effectively identify ADHD children from healthy controls based on fNIRS signals, which argues for the potential utility of fNIRS in future assessments.

Automatic Classification on Multi-Modal MRI Data for Diagnosis of the Postural Instability and Gait Difficulty Subtype of Parkinson's Disease

BACKGROUND

Patients with the postural instability and gait difficulty subtype (PIGD) of Parkinson's disease (PD) are a refractory challenge in clinical practice. Despite previous attempts that have been made at studying subtype-specific brain alterations across PD population, conclusive neuroimaging biomarkers on patients with the PIGD subtype are still lacking. Machine learning-based classifications are a promising tool for differential diagnosis that effectively integrate complex and multivariate data.

OBJECTIVE

Our present study aimed to introduce the machine learning-based automatic classification for the first time to distinguish patients with the PIGD subtype from those with the non-PIGD subtype of PD at the individual level.

METHODS

Fifty-two PD patients and forty-five normal controls (NCs) were recruited and underwent multi-modal MRI scans including a set of resting-state functional, 3D T1-weighted and diffusion tensor imaging sequences. By comparing the PD patients with the NCs, features that were not conducive to the subtype-specific classification were ruled out from massive brain features. We applied a support vector machine classifier with the recursive feature elimination method to multi-modal MRI data for selecting features with the best discriminating power, and evaluated the proposed classifier with the leave-one-out cross-validation.

RESULTS

Using this classifier, we obtained satisfactory diagnostic rates (accuracy = 92.31%, specificity = 96.97%, sensitivity = 84.21% and AUCmax  = 0.9585). The diagnostic agreement evaluated by the Kappa test showed an almost perfect agreement with the existing clinical categorization (Kappa value = 0.83).

CONCLUSIONS

With these favorable results, our findings suggested the machine learning-based classification as an alternative technique to classifying clinical subtypes in PD.

The relevance of feature selection methods to the classification of obsessive-compulsive disorder based on volumetric measures

BACKGROUND

Magnetic resonance images (MRI) show detectable anatomical and functional differences between individuals with obsessive-compulsive disorder (OCD) and healthy subjects. Moreover, machine learning techniques have been proposed as tools to identify potential biomarkers and, ultimately, to support clinical diagnosis. However, few studies to date have investigated feature selection (FS) influences in OCD MRI-based classification.

METHODS

Volumes of cortical and subcortical structures, from MRI data of 38 OCD patients (split into two groups according symptoms severity) and 36 controls, were submitted to seven feature selection algorithms. FS aims to select the most relevant and less redundant features which discriminate between two classes. Then, a classification step was applied, from which the classification performances before and after different FS were compared. For the performance evaluation, leave-one-subject-out accuracies of Support Vector Machine classifiers were considered.

RESULTS

Using different FS algorithms, performance improvement was achieved for Controls vs. All OCD discrimination (19.08% of improvement reducing by 80% the amount of features), Controls vs. Low OCD (20.10%, 75%), Controls vs. High OCD (17.32%, 85%) and Low OCD vs. High OCD (10.53%, 75%). Furthermore, all algorithms pointed out classical cortico-striato-thalamo-cortical circuitry structures as relevant features for OCD classification.

LIMITATIONS

Limitations include the sample size and using only filter approaches for FS.

CONCLUSIONS

Our results suggest that FS positively impacts OCD classification using machine-learning techniques. Complementarily, FS algorithms were able to select biologically plausible features automatically.

Addressing Confounding in Predictive Models with an Application to Neuroimaging

Understanding structural changes in the brain that are caused by a particular disease is a major goal of neuroimaging research. Multivariate pattern analysis (MVPA) comprises a collection of tools that can be used to understand complex disease efxcfects across the brain. We discuss several important issues that must be considered when analyzing data from neuroimaging studies using MVPA. In particular, we focus on the consequences of confounding by non-imaging variables such as age and sex on the results of MVPA. After reviewing current practice to address confounding in neuroimaging studies, we propose an alternative approach based on inverse probability weighting. Although the proposed method is motivated by neuroimaging applications, it is broadly applicable to many problems in machine learning and predictive modeling. We demonstrate the advantages of our approach on simulated and real data examples.

A Computational Model for the Automatic Diagnosis of Attention Deficit Hyperactivity Disorder Based on Functional Brain Volume

In this paper, we investigated the problem of computer-aided diagnosis of Attention Deficit Hyperactivity Disorder (ADHD) using machine learning techniques. With the ADHD-200 dataset, we developed a Support Vector Machine (SVM) model to classify ADHD patients from typically developing controls (TDCs), using the regional brain volumes as predictors. Conventionally, the volume of a brain region was considered to be an anatomical feature and quantified using structural magnetic resonance images. One major contribution of the present study was that we had initially proposed to measure the regional brain volumes using fMRI images. Brain volumes measured from fMRI images were denoted as functional volumes, which quantified the volumes of brain regions that were actually functioning during fMRI imaging. We compared the predictive power of functional volumes with that of regional brain volumes measured from anatomical images, which were denoted as anatomical volumes. The former demonstrated higher discriminative power than the latter for the classification of ADHD patients vs. TDCs. Combined with our two-step feature selection approach which integrated prior knowledge with the recursive feature elimination (RFE) algorithm, our SVM classification model combining functional volumes and demographic characteristics achieved a balanced accuracy of 67.7%, which was 16.1% higher than that of a relevant model published previously in the work of Sato et al. Furthermore, our classifier highlighted 10 brain regions that were most discriminative in distinguishing between ADHD patients and TDCs. These 10 regions were mainly located in occipital lobe, cerebellum posterior lobe, parietal lobe, frontal lobe, and temporal lobe. Our present study using functional images will likely provide new perspectives about the brain regions affected by ADHD.

fNIRS can robustly measure brain activity during memory encoding and retrieval in healthy subjects

Early intervention in Alzheimer’s Disease (AD) requires novel biomarkers that can capture changes in brain activity at an early stage. Current AD biomarkers are expensive and/or invasive and therefore unsuitable for use as screening tools, but a non-invasive, inexpensive, easily accessible screening method could be useful in both clinical and research settings. Prior studies suggest that especially paired-associate learning tasks may be useful in detecting the earliest memory impairment in AD. Here, we investigated the utility of functional Near Infrared Spectroscopy in measuring brain activity from prefrontal, parietal and temporal cortices of healthy adults (n = 19) during memory encoding and retrieval under a face-name paired-associate learning task. Our findings demonstrate that encoding of novel face-name pairs compared to baseline as well as compared to repeated face-name pairs resulted in significant activation in left dorsolateral prefrontal cortex while recalling resulted in activation in dorsolateral prefrontal cortex bilaterally. Moreover, brain response to recalling was significantly higher than encoding in medial, superior and middle frontal cortices for novel faces. Overall, this study shows that fNIRS can reliably measure cortical brain activation during a face-name paired-associate learning task. Future work will include similar measurements in populations with progressing memory deficits.

Reconstruction of Arm Movement Directions from Human Motor Cortex Using fMRI

Recent advances in functional magnetic resonance imaging (fMRI) have been used to reconstruct cognitive states based on brain activity evoked by sensory or cognitive stimuli. To date, such decoding paradigms were mostly used for visual modalities. On the other hand, reconstructing functional brain activity in motor areas was primarily achieved through more invasive electrophysiological techniques. Here, we investigated whether non-invasive fMRI responses from human motor cortex can also be used to predict individual arm movements. To this end, we conducted fMRI studies in which participants moved their arm from a center position to one of eight target directions. Our results suggest that arm movement directions can be distinguished from the multivoxel patterns of fMRI responses in motor cortex. Furthermore, compared to multivoxel pattern analysis, encoding models were able to also reconstruct unknown movement directions from the predicted brain activity. We conclude for our study that non-invasive fMRI signal can be utilized to predict directional motor movements in human motor cortex.

Multivariate Bayesian decoding of single-trial event-related fMRI responses for memory retrieval of voluntary actions

This study proposes a method for classifying event-related fMRI responses in a specialized setting of many known but few unknown stimuli presented in a rapid event-related design. Compared to block design fMRI signals, classification of the response to a single or a few stimulus trial(s) is not a trivial problem due to contamination by preceding events as well as the low signal-to-noise ratio. To overcome such problems, we proposed a single trial-based classification method of rapid event-related fMRI signals utilizing sparse multivariate Bayesian decoding of spatio-temporal fMRI responses. We applied the proposed method to classification of memory retrieval processes for two different classes of episodic memories: a voluntarily conducted experience and a passive experience induced by watching a video of others’ actions. A cross-validation showed higher classification performance of the proposed method compared to that of a support vector machine or of a classifier based on the general linear model. Evaluation of classification performances for one, two, and three stimuli from the same class and a correlation analysis between classification accuracy and target stimulus positions among trials suggest that presenting two target stimuli at longer inter-stimulus intervals is optimal in the design of classification experiments to identify the target stimuli. The proposed method for decoding subject-specific memory retrieval of voluntary behavior using fMRI would be useful in forensic applications in a natural environment, where many known trials can be extracted from a simulation of everyday tasks and few target stimuli from a crime scene.

Identifying current and remitted major depressive disorder with the Hurst exponent: a comparative study on two automated anatomical labeling atlases

Major depressive disorder (MDD) is a leading world-wide psychiatric disorder with high recurrence rate, therefore, it is desirable to identify current MDD (cMDD) and remitted MDD (rMDD) for their appropriate therapeutic interventions. In the study, 19 cMDD, 19 rMDD and 19 well-matched healthy controls (HC) were enrolled and scanned with the resting-state functional magnetic resonance imaging (rs-fMRI). The Hurst exponent (HE) of rs-fMRI in AAL-90 and AAL-1024 atlases were calculated and compared between groups. Then, a radial basis function (RBF) based support vector machine was proposed to identify every pair of the cMDD, rMDD and HC groups using the abnormal HE features, and a leave-one-out cross-validation was used to evaluate the classification performance. Applying the proposed method with AAL-1024 and AAL-90 atlas respectively, 87% and 84% subjects were correctly identified between cMDD and HC, 84% and 71% between rMDD and HC, and 89% and 74% between cMDD and rMDD. Our results indicated that the HE was an effective feature to distinguish cMDD and rMDD from HC, and the recognition performances with AAL-1024 parcellation were better than that with the conventional AAL-90 parcellation.

The Temporal Pattern of a Lesion Modulates the Functional Network Topology of Remote Brain Regions

Focal brain lesions can alter the morphology and function of remote brain areas. When the damage is inflicted more slowly, the functional compensation by and structural reshaping of these areas seem to be more effective. It remains unclear, however, whether the momentum of lesion development also modulates the functional network topology of the remote brain areas. In this study, we compared resting-state functional connectivity data of patients with a slowly growing low-grade glioma (LGG) with that of patients with a faster-growing high-grade glioma (HGG). Using graph theory, we examined whether the tumour growth velocity modulated the functional network topology of remote areas, more specifically of the hemisphere contralateral to the lesion. We observed that the contralesional network topology characteristics differed between patient groups. Based only on the connectivity of the hemisphere contralateral to the lesion, patients could be classified in the correct tumour-grade group with 70% accuracy. Additionally, LGG patients showed smaller contralesional intramodular connectivity, smaller contralesional ratio between intra- and intermodular connectivity, and larger contralesional intermodular connectivity than HGG patients. These results suggest that, in the hemisphere contralateral to the lesion, there is a lower capacity for local, specialized information processing coupled to a higher capacity for distributed information processing in LGG patients. These results underline the utility of a network perspective in evaluating effects of focal brain injury.

Reading-induced shifts of perceptual speech representations in auditory cortex

Learning to read requires the formation of efficient neural associations between written and spoken language. Whether these associations influence the auditory cortical representation of speech remains unknown. Here we address this question by combining multivariate functional MRI analysis and a newly-developed ‘text-based recalibration’ paradigm. In this paradigm, the pairing of visual text and ambiguous speech sounds shifts (i.e. recalibrates) the perceptual interpretation of the ambiguous sounds in subsequent auditory-only trials. We show that it is possible to retrieve the text-induced perceptual interpretation from fMRI activity patterns in the posterior superior temporal cortex. Furthermore, this auditory cortical region showed significant functional connectivity with the inferior parietal lobe (IPL) during the pairing of text with ambiguous speech. Our findings indicate that reading-related audiovisual mappings can adjust the auditory cortical representation of speech in typically reading adults. Additionally, they suggest the involvement of the IPL in audiovisual and/or higher-order perceptual processes leading to this adjustment. When applied in typical and dyslexic readers of different ages, our text-based recalibration paradigm may reveal relevant aspects of perceptual learning and plasticity during successful and failing reading development.

Granger causality reveals a dominant role of memory circuit in chronic opioid dependence

Resting-state magnetic resonance imaging has uncovered abnormal functional connectivity in heroin-dependent individuals (HDIs). However, it remains unclear how brain regions implicated in addictions are related in baseline state without conditioned cues in heroin dependent individuals during opioid maintenance treatment (HDIs-OMT). Previous connectivity analysis assessed the strength of correlated activity between brain regions but lacked the ability to infer directional neural interactions. In the current study, we employed Granger causality analysis to investigate directional causal influences among the brain circuits in HDIs-OMT and non-opioid users. The results revealed a weaker effective connectivity between the caudate nucleus implicated in mediating the reward circuit and other brain regions and also a weaker connectivity between the anterior cingulate cortex and medial prefrontal cortex implicated in mediating inhibitory control. Conversely, HDIs-OMT exhibited stronger effective connectivity between the hippocampus and amygdala implicated in mediating learning-memory, and the anterior cingulate cortex involved in mediating inhibitory control while the putamen mediated learned habits, suggesting that the hippocampus and amygdala may propel the memory circuit to override the control circuit and drive the learned habit in HDIs-OMT. Alterations in learning-memory and inhibitory control may contribute jointly and form a basis for relapse risk even after a period of heroin abstinence. Sustained neural effect of opioid dependence on methadone maintenance including hyperactivation in the memory circuit and impairment in the control circuit support the role of the memory circuitry in relapse and may help redefine targets for treatment.

A multimodal encoding model applied to imaging decision-related neural cascades in the human brain

Perception and cognition in the brain are naturally characterized as spatiotemporal processes. Decision-making, for example, depends on coordinated patterns of neural activity cascading across the brain, running in time from stimulus to response and in space from primary sensory regions to the frontal lobe. Measuring this cascade is key to developing an understanding of brain function. Here we report on a novel methodology that employs multi-modal imaging for inferring this cascade in humans at unprecedented spatiotemporal resolution. Specifically, we develop an encoding model to link simultaneously measured electroencephalography (EEG) and functional magnetic resonance imaging (fMRI) signals to infer high-resolution spatiotemporal brain dynamics during a perceptual decision. After demonstrating replication of results from the literature, we report previously unobserved sequential reactivation of a substantial fraction of the pre-response network whose magnitude correlates with a proxy for decision confidence. Our encoding model, which temporally tags BOLD activations using time localized EEG variability, identifies a coordinated and spatially distributed neural cascade that is associated with a perceptual decision. In general the methodology illuminates complex brain dynamics that would otherwise be unobservable using fMRI or EEG acquired separately.

On memories, neural ensembles and mental flexibility

Memories are assumed to be represented by groups of co-activated neurons, called neural ensembles. Describing ensembles is a challenge: complexity of the underlying micro-circuitry is immense. Current approaches use a piecemeal fashion, focusing on single neurons and employing local measures like pairwise correlations. We introduce an alternative approach that identifies ensembles and describes the effective connectivity between them in a holistic fashion. It also links the oscillatory frequencies observed in ensembles with the spatial scales at which activity is expressed. Using unsupervised learning, biophysical modeling and graph theory, we analyze multi-electrode LFPs from frontal cortex during a spatial delayed response task. We find distinct ensembles for different cues and more parsimonious connectivity for cues on the horizontal axis, which may explain the oblique effect in psychophysics. Our approach paves the way for biophysical models with learned parameters that can guide future Brain Computer Interface development.

Sensitivity to an Illusion of Sound Location in Human Auditory Cortex

Human listeners place greater weight on the beginning of a sound compared to the middle or end when determining sound location, creating an auditory illusion known as the Franssen effect. Here, we exploited that effect to test whether human auditory cortex (AC) represents the physical vs. perceived spatial features of a sound. We used functional magnetic resonance imaging (fMRI) to measure AC responses to sounds that varied in perceived location due to interaural level differences (ILD) applied to sound onsets or to the full sound duration. Analysis of hemodynamic responses in AC revealed sensitivity to ILD in both full-cue (veridical) and onset-only (illusory) lateralized stimuli. Classification analysis revealed regional differences in the sensitivity to onset-only ILDs, where better classification was observed in posterior compared to primary AC. That is, restricting the ILD to sound onset—which alters the physical but not the perceptual nature of the spatial cue—did not eliminate cortical sensitivity to that cue. These results suggest that perceptual representations of auditory space emerge or are refined in higher-order AC regions, supporting the stable perception of auditory space in noisy or reverberant environments and forming the basis of illusions such as the Franssen effect.

Object Recognition in Mental Representations: Directions for Exploring Diagnostic Features through Visual Mental Imagery

One of the fundamental goals of object recognition research is to understand how a cognitive representation produced from the output of filtered and transformed sensory information facilitates efficient viewer behavior. Given that mental imagery strongly resembles perceptual processes in both cortical regions and subjective visual qualities, it is reasonable to question whether mental imagery facilitates cognition in a manner similar to that of perceptual viewing: via the detection and recognition of distinguishing features. Categorizing the feature content of mental imagery holds potential as a reverse pathway by which to identify the components of a visual stimulus which are most critical for the creation and retrieval of a visual representation. This review will examine the likelihood that the information represented in visual mental imagery reflects distinctive object features thought to facilitate efficient object categorization and recognition during perceptual viewing. If it is the case that these representational features resemble their sensory counterparts in both spatial and semantic qualities, they may well be accessible through mental imagery as evaluated through current investigative techniques. In this review, methods applied to mental imagery research and their findings are reviewed and evaluated for their efficiency in accessing internal representations, and implications for identifying diagnostic features are discussed. An argument is made for the benefits of combining mental imagery assessment methods with diagnostic feature research to advance the understanding of visual perceptive processes, with suggestions for avenues of future investigation.

The impact of ultra-high field MRI on cognitive and computational neuroimaging

The ability to measure functional brain responses non-invasively with ultra high field MRI (7 T and above) represents a unique opportunity in advancing our understanding of the human brain. Compared to lower fields (3 T and below), ultra high field MRI has an increased sensitivity, which can be used to acquire functional images with greater spatial resolution, and greater specificity of the blood oxygen level dependent (BOLD) signal to the underlying neuronal responses. Together, increased resolution and specificity enable investigating brain functions at a submillimeter scale, which so far could only be done with invasive techniques. At this mesoscopic spatial scale, perception, cognition and behavior can be probed at the level of fundamental units of neural computations, such as cortical columns, cortical layers, and subcortical nuclei. This represents a unique and distinctive advantage that differentiates ultra high from lower field imaging and that can foster a tighter link between fMRI and computational modeling of neural networks. So far, functional brain mapping at submillimeter scale has focused on the processing of sensory information and on well-known systems for which extensive information is available from invasive recordings in animals. It remains an open challenge to extend this methodology to uniquely human functions and, more generally, to systems for which animal models may be problematic. To succeed, the possibility to acquire high-resolution functional data with large spatial coverage, the availability of computational models of neural processing as well as accurate biophysical modeling of neurovascular coupling at mesoscopic scale all appear necessary.

Decoding Dynamic Brain Patterns from Evoked Responses: A Tutorial on Multivariate Pattern Analysis Applied to Time Series Neuroimaging Data

Multivariate pattern analysis (MVPA) or brain decoding methods have become standard practice in analyzing fMRI data. Although decoding methods have been extensively applied in brain–computer interfaces, these methods have only recently been applied to time series neuroimaging data such as MEG and EEG to address experimental questions in cognitive neuroscience. In a tutorial style review, we describe a broad set of options to inform future time series decoding studies from a cognitive neuroscience perspective. Using example MEG data, we illustrate the effects that different options in the decoding analysis pipeline can have on experimental results where the aim is to “decode” different perceptual stimuli or cognitive states over time from dynamic brain activation patterns. We show that decisions made at both preprocessing (e.g., dimensionality reduction, subsampling, trial averaging) and decoding (e.g., classifier selection, cross-validation design) stages of the analysis can significantly affect the results. In addition to standard decoding, we describe extensions to MVPA for time-varying neuroimaging data including representational similarity analysis, temporal generalization, and the interpretation of classifier weight maps. Finally, we outline important caveats in the design and interpretation of time series decoding experiments.

Altered Topological Properties of Brain Networks in Social Anxiety Disorder: A Resting-state Functional MRI Study

Recent studies involving connectome analysis including graph theory have yielded potential biomarkers for mental disorders. In this study, we aimed to investigate the differences of resting-state network between patients with social anxiety disorder (SAD) and healthy controls (HCs), as well as to distinguish between individual subjects using topological properties. In total, 42 SAD patients and the same number of HCs underwent resting functional MRI, and the topological organization of the whole-brain functional network was calculated using graph theory. Compared with the controls, the patients showed a decrease in 49 positive connections. In the topological analysis, the patients showed an increase in the area under the curve (AUC) of the global shortest path length of the network (L_p) and a decrease in the AUC of the global clustering coefficient of the network (C_p). Furthermore, the AUCs of L_p and C_p were used to effectively discriminate the individual SAD patients from the HCs with high accuracy. This study revealed that the neural networks of the SAD patients showed changes in topological characteristics, and these changes were prominent not only in both groups but also at the individual level. This study provides a new perspective for the identification of patients with SAD.

Multiregional integration in the brain during resting-state fMRI activity

Data-driven models of functional magnetic resonance imaging (fMRI) activity can elucidate dependencies that involve the combination of multiple brain regions. Activity in some regions during resting-state fMRI can be predicted with high accuracy from the activities of other regions. However, it remains unclear in which regions activity depends on unique integration of multiple predictor regions. To address this question, sparse (parsimonious) models could serve to better determine key interregional dependencies by reducing false positives. We used resting-state fMRI data from 46 subjects, and for each region of interest (ROI) per subject we performed whole-brain recursive feature elimination (RFE) to select the minimal set of ROIs that best predicted activity in the modeled ROI. We quantified the dependence of activity on multiple predictor ROIs, by measuring the gain in prediction accuracy of models that incorporated multiple predictor ROIs compared to models that used a single predictor ROI. We identified regions that showed considerable evidence of multiregional integration and determined the key regions that contributed to their observed activity. Our models reveal fronto-parietal integration networks, little integration in primary sensory regions, as well as redundancy between some regions. Our study demonstrates the utility of whole-brain RFE to generate data-driven models with minimal sets of ROIs that predict activity with high accuracy. By determining the extent to which activity in each ROI depended on integration of signals from multiple ROIs, we find cortical integration networks during resting-state activity.

Models of fMRI activity can elucidate underlying dependencies that involve the combination of multiple brain regions. However, it remains unclear in which regions activity depends on unique integration of multiple predictor regions. To address this question, sparse (parsimonious) models could serve to better determine key interregional dependencies by reducing false positives. We used resting-state fMRI data, and for each brain region we performed whole-brain recursive feature elimination to select the minimal set of regions that best predicted activity in the region. We identified integrator regions by quantifying the gain in prediction accuracy of models that incorporated multiple predictor regions compared to single predictor region. Our study provides data-driven models that use minimal sets of regions to predict activity with high accuracy. By determining the extent to which activity in each region depended on integration of signals from multiple sources, we find cortical integration networks during resting-state activity.

Improved individualized prediction of schizophrenia in subjects at familial high risk, based on neuroanatomical data, schizotypal and neurocognitive features

To date, there are no reliable markers for predicting onset of schizophrenia in individuals at high risk (HR). Substantial promise is, however, shown by a variety of pattern classification approaches to neuroimaging data. Here, we examined the predictive accuracy of support vector machine (SVM) in later diagnosing schizophrenia, at a single-subject level, using a cohort of HR individuals drawn from multiply affected families and a combination of neuroanatomical, schizotypal and neurocognitive variables. Baseline structural magnetic resonance imaging (MRI), schizotypal and neurocognitive data from 17 HR subjects, who subsequently developed schizophrenia and a matched group of 17 HR subjects who did not make the transition, yet had psychotic symptoms, were included in the analysis. We employed recursive feature elimination (RFE), in a nested cross-validation scheme to identify the most significant predictors of disease transition and enhance diagnostic performance. Classification accuracy was 94% when a self-completed measure of schizotypy, a declarative memory test and structural MRI data were combined into a single learning algorithm; higher than when either quantitative measure was used alone. The discriminative neuroanatomical pattern involved gray matter volume differences in frontal, orbito-frontal and occipital lobe regions bilaterally as well as parts of the superior, medial temporal lobe and cerebellar regions. Our findings suggest that an early SVM-based prediction of schizophrenia is possible and can be improved by combining schizotypal and neurocognitive features with neuroanatomical variables. However, our predictive model needs to be tested by classifying a new, independent HR cohort in order to estimate its validity.

Brief Mental Training Reorganizes Large-Scale Brain Networks

Emerging evidences have shown that one form of mental training—mindfulness meditation, can improve attention, emotion regulation and cognitive performance through changing brain activity and structural connectivity. However, whether and how the short-term mindfulness meditation alters large-scale brain networks are not well understood. Here, we applied a novel data-driven technique, the multivariate pattern analysis (MVPA) to resting-state fMRI (rsfMRI) data to identify changes in brain activity patterns and assess the neural mechanisms induced by a brief mindfulness training—integrative body–mind training (IBMT), which was previously reported in our series of randomized studies. Whole brain rsfMRI was performed on an undergraduate group who received 2 weeks of IBMT with 30 min per session (5 h training in total). Classifiers were trained on measures of functional connectivity in this fMRI data, and they were able to reliably differentiate (with 72% accuracy) patterns of connectivity from before vs. after the IBMT training. After training, an increase in positive functional connections (60 connections) were detected, primarily involving bilateral superior/middle occipital gyrus, bilateral frontale operculum, bilateral superior temporal gyrus, right superior temporal pole, bilateral insula, caudate and cerebellum. These results suggest that brief mental training alters the functional connectivity of large-scale brain networks at rest that may involve a portion of the neural circuitry supporting attention, cognitive and affective processing, awareness and sensory integration and reward processing.

Functional connectomics from a "big data" perspective

In the last decade, explosive growth regarding functional connectome studies has been observed. Accumulating knowledge has significantly contributed to our understanding of the brain's functional network architectures in health and disease. With the development of innovative neuroimaging techniques, the establishment of large brain datasets and the increasing accumulation of published findings, functional connectomic research has begun to move into the era of "big data", which generates unprecedented opportunities for discovery in brain science and simultaneously encounters various challenging issues, such as data acquisition, management and analyses. Big data on the functional connectome exhibits several critical features: high spatial and/or temporal precision, large sample sizes, long-term recording of brain activity, multidimensional biological variables (e.g., imaging, genetic, demographic, cognitive and clinic) and/or vast quantities of existing findings. We review studies regarding functional connectomics from a big data perspective, with a focus on recent methodological advances in state-of-the-art image acquisition (e.g., multiband imaging), analysis approaches and statistical strategies (e.g., graph theoretical analysis, dynamic network analysis, independent component analysis, multivariate pattern analysis and machine learning), as well as reliability and reproducibility validations. We highlight the novel findings in the application of functional connectomic big data to the exploration of the biological mechanisms of cognitive functions, normal development and aging and of neurological and psychiatric disorders. We advocate the urgent need to expand efforts directed at the methodological challenges and discuss the direction of applications in this field.

Degradation of cortical representations during encoding following sleep deprivation

A night of total sleep deprivation (TSD) reduces task-related activation of fronto-parietal and higher visual cortical areas. As this reduction in activation corresponds to impaired attention and perceptual processing, it might also be associated with poorer memory encoding. Related animal work has established that cortical columns stochastically enter a 'down' state in sleep deprivation, leading to predictions that neural representations are less stable and distinctive following TSD. To test these predictions participants incidentally encoded scene images while undergoing fMRI, either during rested wakefulness (RW) or after TSD. In scene-selective PPA, TSD reduced stability of neural representations across repetition. This was accompanied by poorer subsequent memory. Greater representational stability benefitted subsequent memory in RW but not TSD. Even for items subsequently recognized, representational distinctiveness was lower in TSD, suggesting that quality of encoding is degraded. Reduced representational stability and distinctiveness are two novel mechanisms by which TSD can contribute to poorer memory formation.

Representations of Pitch and Timbre Variation in Human Auditory Cortex

Pitch and timbre are two primary dimensions of auditory perception, but how they are represented in the human brain remains a matter of contention. Some animal studies of auditory cortical processing have suggested modular processing, with different brain regions preferentially coding for pitch or timbre, whereas other studies have suggested a distributed code for different attributes across the same population of neurons. This study tested whether variations in pitch and timbre elicit activity in distinct regions of the human temporal lobes. Listeners were presented with sequences of sounds that varied in either fundamental frequency (eliciting changes in pitch) or spectral centroid (eliciting changes in brightness, an important attribute of timbre), with the degree of pitch or timbre variation in each sequence parametrically manipulated. The BOLD responses from auditory cortex increased with increasing sequence variance along each perceptual dimension. The spatial extent, region, and laterality of the cortical regions most responsive to variations in pitch or timbre at the univariate level of analysis were largely overlapping. However, patterns of activation in response to pitch or timbre variations were discriminable in most subjects at an individual level using multivoxel pattern analysis, suggesting a distributed coding of the two dimensions bilaterally in human auditory cortex.

SIGNIFICANCE STATEMENT

Pitch and timbre are two crucial aspects of auditory perception. Pitch governs our perception of musical melodies and harmonies, and conveys both prosodic and (in tone languages) lexical information in speech. Brightness-an aspect of timbre or sound quality-allows us to distinguish different musical instruments and speech sounds. Frequency-mapping studies have revealed tonotopic organization in primary auditory cortex, but the use of pure tones or noise bands has precluded the possibility of dissociating pitch from brightness. Our results suggest a distributed code, with no clear anatomical distinctions between auditory cortical regions responsive to changes in either pitch or timbre, but also reveal a population code that can differentiate between changes in either dimension within the same cortical regions.

What the success of brain imaging implies about the neural code

The success of fMRI places constraints on the nature of the neural code. The fact that researchers can infer similarities between neural representations, despite fMRI’s limitations, implies that certain neural coding schemes are more likely than others. For fMRI to succeed given its low temporal and spatial resolution, the neural code must be smooth at the voxel and functional level such that similar stimuli engender similar internal representations. Through proof and simulation, we determine which coding schemes are plausible given both fMRI’s successes and its limitations in measuring neural activity. Deep neural network approaches, which have been forwarded as computational accounts of the ventral stream, are consistent with the success of fMRI, though functional smoothness breaks down in the later network layers. These results have implications for the nature of the neural code and ventral stream, as well as what can be successfully investigated with fMRI.

DOI:http://dx.doi.org/10.7554/eLife.21397.001

We can appreciate that a cat is more similar to a dog than to a truck. The combined activity of millions of neurons in the brain somehow captures these everyday similarities, and this activity can be measured using imaging techniques such as functional magnetic resonance imaging (fMRI). However, fMRI scanners are not particularly precise – they average together the responses of many thousands of neurons over several seconds, which provides a blurry snapshot of brain activity. Nevertheless, the pattern of activity measured when viewing a photograph of a cat is more similar to that seen when viewing a picture of a dog than a picture of a truck. This tells us a lot about how the brain codes information, as only certain coding methods would allow fMRI to capture these similarities given the technique’s limitations.

There are many different models that attempt to describe how the brain codes similarity relations. Some models use the principle of neural networks, in which neurons can be considered as arranged into interconnected layers. In such models, neurons transmit information from one layer to the next.

By investigating which models are consistent with fMRI’s ability to capture similarity relations, Guest and Love have found that certain neural network models are plausible accounts of how the brain represents and processes information. These models include the deep learning networks that contain many layers of neurons and are popularly used in artificial intelligence. Other modeling approaches do not account for the ability of fMRI to capture similarity relations.

As neural networks become deeper with more layers, they should be less readily understood using fMRI: as the number of layers increases, the representations of objects with similarities (for example, cats and dogs) become more unrelated. One question that requires further investigation is whether this finding explains why certain parts of the brain are more difficult to image.

DOI:http://dx.doi.org/10.7554/eLife.21397.002

Nicotine deprivation elevates neural representation of smoking-related cues in object-sensitive visual cortex: a proof of concept study

OBJECTIVE

In the current study, we use functional magnetic resonance imaging (fMRI) and multi-voxel pattern analysis (MVPA) to investigate whether tobacco addiction biases basic visual processing in favour of smoking-related images. We hypothesize that the neural representation of smoking-related stimuli in the lateral occipital complex (LOC) is elevated after a period of nicotine deprivation compared to a satiated state, but that this is not the case for object categories unrelated to smoking.

METHODS

Current smokers (≥10 cigarettes a day) underwent two fMRI scanning sessions: one after 10 h of nicotine abstinence and the other one after smoking ad libitum. Regional blood oxygenated level-dependent (BOLD) response was measured while participants were presented with 24 blocks of 8 colour-matched pictures of cigarettes, pencils or chairs. The functional data of 10 participants were analysed through a pattern classification approach.

RESULTS

In bilateral LOC clusters, the classifier was able to discriminate between patterns of activity elicited by visually similar smoking-related (cigarettes) and neutral objects (pencils) above empirically estimated chance levels only during deprivation (mean = 61.0%, chance (permutations) = 50.0%, p = .01) but not during satiation (mean = 53.5%, chance (permutations) = 49.9%, ns.). For all other stimulus contrasts, there was no difference in discriminability between the deprived and satiated conditions.

CONCLUSION

The discriminability between smoking and non-smoking visual objects was elevated in object-selective brain region LOC after a period of nicotine abstinence. This indicates that attention bias likely affects basic visual object processing.

Integration of Bioelectronics and Bioinformatics: Future Direction of Bioengineering Research

This special issue of the Journal of Medical and Biological Engineering highlights the field of advanced bioelectronics and bioinformatics. Several papers were considered for this special issue, including those on bioelectronics in wearable and implantable medical devices, such as sensors, and bioinformatics in healthcare, brain cognition, and various neural pathologies. Many investigators contributed original research articles to this issue, demonstrating emerging research fields. More than 20 papers were accepted for publication after a high-quality critical review was conducted, and 14 papers were selected for this special issue. This special issue on bioelectronics and bioinformatics attracted a substantial number of full-paper submissions from many countries. We appreciate the numerous volunteers who helped review the manuscripts. This paper provides a brief review of issues regarding bioelectronics and bioinformatics devices.

Promises, Pitfalls, and Basic Guidelines for Applying Machine Learning Classifiers to Psychiatric Imaging Data, with Autism as an Example

Most psychiatric disorders are associated with subtle alterations in brain function and are subject to large interindividual differences. Typically, the diagnosis of these disorders requires time-consuming behavioral assessments administered by a multidisciplinary team with extensive experience. While the application of Machine Learning classification methods (ML classifiers) to neuroimaging data has the potential to speed and simplify diagnosis of psychiatric disorders, the methods, assumptions, and analytical steps are currently opaque and not accessible to researchers and clinicians outside the field. In this paper, we describe potential classification pipelines for autism spectrum disorder, as an example of a psychiatric disorder. The analyses are based on resting-state fMRI data derived from a multisite data repository (ABIDE). We compare several popular ML classifiers such as support vector machines, neural networks, and regression approaches, among others. In a tutorial style, written to be equally accessible for researchers and clinicians, we explain the rationale of each classification approach, clarify the underlying assumptions, and discuss possible pitfalls and challenges. We also provide the data as well as the MATLAB code we used to achieve our results. We show that out-of-the-box ML classifiers can yield classification accuracies of about 60-70%. Finally, we discuss how classification accuracy can be further improved, and we mention methodological developments that are needed to pave the way for the use of ML classifiers in clinical practice.

"Hearing faces and seeing voices": Amodal coding of person identity in the human brain

Recognizing familiar individuals is achieved by the brain by combining cues from several sensory modalities, including the face of a person and her voice. Here we used functional magnetic resonance (fMRI) and a whole-brain, searchlight multi-voxel pattern analysis (MVPA) to search for areas in which local fMRI patterns could result in identity classification as a function of sensory modality. We found several areas supporting face or voice stimulus classification based on fMRI responses, consistent with previous reports; the classification maps overlapped across modalities in a single area of right posterior superior temporal sulcus (pSTS). Remarkably, we also found several cortical areas, mostly located along the middle temporal gyrus, in which local fMRI patterns resulted in identity “cross-classification”: vocal identity could be classified based on fMRI responses to the faces, or the reverse, or both. These findings are suggestive of a series of cortical identity representations increasingly abstracted from the input modality.

Pairwise Classifier Ensemble with Adaptive Sub-Classifiers for fMRI Pattern Analysis

The multi-voxel pattern analysis technique is applied to fMRI data for classification of high-level brain functions using pattern information distributed over multiple voxels. In this paper, we propose a classifier ensemble for multiclass classification in fMRI analysis, exploiting the fact that specific neighboring voxels can contain spatial pattern information. The proposed method converts the multiclass classification to a pairwise classifier ensemble, and each pairwise classifier consists of multiple sub-classifiers using an adaptive feature set for each class-pair. Simulated and real fMRI data were used to verify the proposed method. Intra- and inter-subject analyses were performed to compare the proposed method with several well-known classifiers, including single and ensemble classifiers. The comparison results showed that the proposed method can be generally applied to multiclass classification in both simulations and real fMRI analyses.

Patterns of Activity in the Human Frontal and Parietal Cortex Differentiate Large and Small Saccades

A vast literature indicates that small and large saccades, respectively, subserve different perceptual and cognitive strategies and may rely on different programming modes. While it is well-established that in monkeys’ main oculomotor brain regions small and large eye movements are controlled by segregated neuronal populations, the representation of saccade amplitude in the human brain remains unclear. To address this question we used functional magnetic resonance imaging to scan participants while they performed saccades toward targets at either short (4°) or large (30°) eccentricity. A regional multivoxel pattern analysis reveals that patterns of activity in the frontal eye-field and parietal eye fields discriminate between the execution of large or small saccades. This was not the case in the supplementary eye-fields nor in the inferior precentral cortex. These findings provide the first evidence of a representation of saccadic eye movement size in the fronto-parietal occulomotor circuit. They shed light on the respective roles of the different cortical oculomotor regions with respect to space perception and exploration, as well as on the homology of eye movement control between human and non-human primates.

Anomalous gray matter patterns in specific reading comprehension deficit are independent of dyslexia

Specific reading comprehension deficit (SRCD) affects up to 10 % of all children. SRCD is distinct from dyslexia (DYS) in that individuals with SRCD show poor comprehension despite adequate decoding skills. Despite its prevalence and considerable behavioral research, there is not yet a unified cognitive profile of SRCD. While its neuroanatomical basis is unknown, SRCD could be anomalous in regions subserving their commonly reported cognitive weaknesses in semantic processing or executive function. Here we investigated, for the first time, patterns of gray matter volume difference in SRCD as compared to DYS and typical developing (TD) adolescent readers (N = 41). A linear support vector machine algorithm was applied to whole brain gray matter volumes generated through voxel-based morphometry. As expected, DYS differed significantly from TD in a pattern that included features from left fusiform and supramarginal gyri (DYS vs. TD: 80.0 %, p < 0.01). SRCD was well differentiated not only from TD (92.5 %, p < 0.001) but also from DYS (88.0 %, p < 0.001). Of particular interest were findings of reduced gray matter volume in right frontal areas that were also supported by univariate analysis. These areas are thought to subserve executive processes relevant for reading, such as monitoring and manipulating mental representations. Thus, preliminary analyses suggest that SRCD readers possess a distinct neural profile compared to both TD and DYS readers and that these differences might be linked to domain-general abilities. This work provides a foundation for further investigation into variants of reading disability beyond DYS.

Select and Cluster: A Method for Finding Functional Networks of Clustered Voxels in fMRI

Extracting functional connectivity patterns among cortical regions in fMRI datasets is a challenge stimulating the development of effective data-driven or model based techniques. Here, we present a novel data-driven method for the extraction of significantly connected functional ROIs directly from the preprocessed fMRI data without relying on a priori knowledge of the expected activations. This method finds spatially compact groups of voxels which show a homogeneous pattern of significant connectivity with other regions in the brain. The method, called Select and Cluster (S&C), consists of two steps: first, a dimensionality reduction step based on a blind multiresolution pairwise correlation by which the subset of all cortical voxels with significant mutual correlation is selected and the second step in which the selected voxels are grouped into spatially compact and functionally homogeneous ROIs by means of a Support Vector Clustering (SVC) algorithm. The S&C method is described in detail. Its performance assessed on simulated and experimental fMRI data is compared to other methods commonly used in functional connectivity analyses, such as Independent Component Analysis (ICA) or clustering. S&C method simplifies the extraction of functional networks in fMRI by identifying automatically spatially compact groups of voxels (ROIs) involved in whole brain scale activation networks.