Predicting hemispheric dominance for language production in healthy individuals using support vector machine

Imaging genetics of language network functional connectivity reveals links with language-related abilities, dyslexia and handedness

Language is supported by a distributed network of brain regions with a particular contribution from the left hemisphere. A multi-level understanding of this network requires studying its genetic architecture. We used resting-state imaging data from 29,681 participants (UK Biobank) to measure connectivity between 18 left-hemisphere regions involved in multimodal sentence-level processing, as well as their right-hemisphere homotopes, and interhemispheric connections. Multivariate genome-wide association analysis of this total network, based on genetic variants with population frequencies  >1%, identified 14 genomic loci, of which three were also associated with asymmetry of intrahemispheric connectivity. Polygenic dispositions to lower language-related abilities, dyslexia and left-handedness were associated with generally reduced leftward asymmetry of functional connectivity. Exome-wide association analysis based on rare, protein-altering variants (frequencies <1%) suggested 7 additional genes. These findings shed new light on genetic contributions to language network organization and related behavioural traits.

Impact of cannabis use on brain metabolism using 31P and 1H magnetic resonance spectroscopy

PURPOSE

This prospective cross-sectional study investigated the influence of regular cannabis use on brain metabolism in young cannabis users by using combined proton and phosphorus magnetic resonance spectroscopy.

METHODS

The study was performed in 45 young cannabis users aged 18-30, who had been using cannabis on a regular basis over a period of at least 2 years and in 47 age-matched controls. We acquired 31P MRS data in different brain regions at 3T with a double-resonant 1H/31P head coil, anatomic images, and 1H MRS data with a standard 20-channel 1H head coil. Absolute concentration values of proton metabolites were obtained via calibration from tissue water as an internal reference, whereas a standard solution of 75 mmol/l KH2PO4 was used as an external reference for the calibration of phosphorus signals.

RESULTS

We found an overall but not statistically significant lower concentration level of several proton and phosphorus metabolites in cannabis users compared to non-users. In particular, energy-related phosphates such as adenosine triphosphate (ATP) and inorganic phosphate (Pi) were reduced in all regions under investigation. Phosphocreatine (PCr) showed lowered values mainly in the left basal ganglia and the left frontal white matter.

CONCLUSION

The results suggest that the increased risk of functional brain disorders observed in long-term cannabis users could be caused by an impairment of the energy metabolism of the brain, but this needs to be verified in future studies.

Spatial Variation Generation Algorithm for Motor Imagery Data Augmentation: Increasing the Density of Sample Vicinity

The imbalanced development between deep learning-based model design and motor imagery (MI) data acquisition raises concerns about the potential overfitting issue-models can identify training data well but fail to generalize test data. In this study, a Spatial Variation Generation (SVG) algorithm for MI data augmentation is proposed to alleviate the overfitting issue. In essence, SVG generates MI data using variations of electrode placement and brain spatial pattern, ultimately elevating the density of the raw sample vicinity. The proposed SVG prevents models from memorizing the training data by replacing the raw samples with the proper vicinal distribution. Moreover, SVG generates a uniform distribution and stabilizes the training process of models. In comparison studies involving five deep learning-based models across eight datasets, the proposed SVG algorithm exhibited a notable improvement of 0.021 in the area under the receiver operating characteristic curve (AUC). The improvement achieved by SVG outperforms other data augmentation algorithms. Further results from the ablation study verify the effectiveness of each component of SVG. Finally, the studies in the control group with varying numbers of samples show that the SVG algorithm consistently improves the AUC, with improvements ranging from approximately 0.02 to 0.15.

Functional MRI in Neuro-Oncology: State of the Art and Future Directions

Since its discovery in the early 1990s, functional MRI (fMRI) has been used to study human brain function. One well-established application of fMRI in the clinical setting is the neurosurgical planning of patients with brain tumors near eloquent cortical areas. Clinical fMRI aims to preoperatively identify eloquent cortices that serve essential functions in daily life, such as hand movement and language. The primary goal of neurosurgery is to maximize tumor resection while sparing eloquent cortices adjacent to the tumor. When a lesion presents in the vicinity of an eloquent cortex, surgeons may use fMRI to plan their best surgical approach by determining the proximity of the lesion to regions of activation, providing guidance for awake brain surgery and intraoperative brain mapping. The acquisition of fMRI requires patient preparation prior to imaging, determination of functional paradigms, monitoring of patient performance, and both processing and analysis of images. Interpretation of fMRI maps requires a strong understanding of functional neuroanatomy and familiarity with the technical limitations frequently present in brain tumor imaging, including neurovascular uncoupling, patient compliance, and data analysis. This review discusses clinical fMRI in neuro-oncology, relevant ongoing research topics, and prospective future developments in this exciting discipline.

Atypical Brain Asymmetry in Human Situs Inversus: Gut Feeling or Real Evidence?

The alignment of visceral and brain asymmetry observed in some vertebrate species raises the question of whether this association also exists in humans. While the visceral and brain systems may have developed asymmetry for different reasons, basic visceral left–right differentiation mechanisms could have been duplicated to establish brain asymmetry. We describe the main phenotypical anomalies and the general mechanism of left–right differentiation of vertebrate visceral and brain laterality. Next, we systematically review the available human studies that explored the prevalence of atypical behavioral and brain asymmetry in visceral situs anomalies, which almost exclusively involved participants with the mirrored visceral organization (situs inversus). The data show no direct link between human visceral and brain functional laterality as most participants with situs inversus show the typical population bias for handedness and brain functional asymmetry, although an increased prevalence of functional crowding may be present. At the same time, several independent studies present evidence for a possible relation between situs inversus and the gross morphological asymmetry of the brain torque with potential differences between subtypes of situs inversus with ciliary and non-ciliary etiologies.

Relations between hemispheric asymmetries of grey matter and auditory processing of spoken syllables in 281 healthy adults

Most people have a right-ear advantage for the perception of spoken syllables, consistent with left hemisphere dominance for speech processing. However, there is considerable variation, with some people showing left-ear advantage. The extent to which this variation is reflected in brain structure remains unclear. We tested for relations between hemispheric asymmetries of auditory processing and of grey matter in 281 adults, using dichotic listening and voxel-based morphometry. This was the largest study of this issue to date. Per-voxel asymmetry indexes were derived for each participant following registration of brain magnetic resonance images to a template that was symmetrized. The asymmetry index derived from dichotic listening was related to grey matter asymmetry in clusters of voxels corresponding to the amygdala and cerebellum lobule VI. There was also a smaller, non-significant cluster in the posterior superior temporal gyrus, a region of auditory cortex. These findings contribute to the mapping of asymmetrical structure–function links in the human brain and suggest that subcortical structures should be investigated in relation to hemispheric dominance for speech processing, in addition to auditory cortex.

Typical and atypical language brain organization based on intrinsic connectivity and multitask functional asymmetries

Based on the joint investigation in 287 healthy volunteers (150 left-Handers (LH)) of language task-induced asymmetries and intrinsic connectivity strength of the sentence-processing supramodal network, we show that individuals with atypical rightward language lateralization (N = 30, 25 LH) do not rely on an organization that simply mirrors that of typical leftward lateralized individuals. Actually, the resting-state organization in the atypicals showed that their sentence processing was underpinned by left and right networks both wired for language processing and highly interacting by strong interhemispheric intrinsic connectivity and larger corpus callosum volume. Such a loose hemispheric specialization for language permits the hosting of language in either the left and/or right hemisphere as assessed by a very high incidence of dissociations across various language task-induced asymmetries in this group.

Left-handed musicians show a higher probability of atypical cerebral dominance for language

Music processing and right hemispheric language lateralization share a common network in the right auditory cortex and its frontal connections. Given that the development of hemispheric language dominance takes place over several years, this study tested whether musicianship could increase the probability of observing right language dominance in left-handers. Using a classic fMRI language paradigm, results showed that atypical lateralization was more predominant in musicians (40%) than in nonmusicians (5%). Comparison of left-handers with typical left and atypical right lateralization revealed that: (a) atypical cases presented a thicker right pars triangularis and more gyrified left Heschl's gyrus; and (b) the right pars triangularis of atypical cases showed a stronger intra-hemispheric functional connectivity with the right angular gyrus, but a weaker interhemispheric functional connectivity with part of the left Broca's area. Thus, musicianship is the first known factor related to a higher prevalence of atypical language dominance in healthy left-handed individuals. We suggest that differences in the frontal and temporal cortex might act as shared predisposing factors to both musicianship and atypical language lateralization.

Stage detection of mild cognitive impairment via fMRI using Hilbert Huang transform based classification framework

PURPOSE

This work aims to establish a classification framework for the diagnosis of mild cognitive impairment (MCI) at different stages (early MCI and late MCI) through direct analysis of resting-state functional magnetic resonance imaging (rs-fMRI) signals and using the accuracy (total correct rate), specificity (correct rate of late MCI) and sensitivity (correct rate of early MCI) to validate its classification performance.

METHODS

All fMR images of subjects were parcellated into 116 regions of interest (ROIs) by applying the Anatomical Automatic Labeling (AAL) template, and the average rs-fMRI signals of each ROI were extracted. The Hilbert-Huang transform (HHT) was introduced into the framework to decompose each rs-fMRI signal into a series of intrinsic mode functions (IMFs) and to analyze these nonstationary and nonlinear time-series from the perspective of multiresolution. After obtaining the instantaneous frequencies and amplitudes of all IMFs of a signal, the Hilbert weighted frequencies (HWFs) were calculated and combined into a vector as the feature of the corresponding ROI. Support Vector Machine (SVM) was implemented to classify MCI at different stages. We used the independent two-sample t-test as the feature selection method and measured the classification performance through the leave-one-out cross-validation (LOOCV) method.

RESULTS

Results on 77 early MCI (eMCI) and 64 late MCI (lMCI) with baseline rs-fMRI data from the Alzheimer's Disease Neuroimaging Initiative (ADNI) yielded 87.94% classification accuracy. Some of the brain regions with significant differences found by previous studies have been confirmed in this work. We found that HWF characteristics exhibited a significant downward trend in all cerebellar regions. The rs-fMRI signals in differential brain regions have not changed completely, but only altered in some narrow frequency bands. The analysis results showed that during the progress of MCI, the main changes of rs-fMRI were concentrated in IMF3, while IMFs with other indexes also contained HWF features with high SVM weights, such as Orbitofrontal superior frontal gyrus in IMF2, Insula in IMF4, and Lobule Ⅲ of vermis in IMF5, indicating that other IMFs provide important information for the diagnosis of MCI as well.

CONCLUSIONS

This work confirmed the classification ability of HHT-based classification framework in classification of at different stages of MCI. Through the analysis, we found that during the progress of MCI the main changes of rs-fMRI were concentrated in IMF3, and HWF characteristics showed a significant downward trend in all cerebellar regions.

Development of handedness, anatomical and functional brain lateralization

The present chapter offers a report on the recent literature on the neural bases of hemispheric specialization (HS), anatomical and functional developmental timecourse of HS, and on the available knowledge of their relationships with the development of handedness. Strong anatomical asymmetries can be seen located along the end of the sylvian fissure and the superior temporal sulcus (STS) as soon as the 23rd gestational week. They correspond to a leftward sulcal depth asymmetry of the Sylvian fissure coupled with a rightward asymmetry of STS. These neonatal asymmetries targeting speech processing areas do not further change with development. Different from these anatomical asymmetries, the functional asymmetries of language areas develop during childhood. Such a development is characterized at birth by a predominant interhemispheric intrinsic connectivity between homotopic areas that will evolve toward left hemisphere intrahemispheric intrinsic connectivity between anterior and posterior language poles. The development of such a typical architecture of language networks in the left hemisphere dominant for language in more than 90% of humans translates into a continuous increase in the leftward asymmetries of activation during language production throughout childhood. With regard to the rightward cerebral lateralization for visuospatial functions, neuroimaging studies tend to indicate an increase in rightward lateralization of frontal-parietal network with age during visuospatial memory and visuospatial search tasks. In addition, the spatial-attentional behavioral asymmetries emerge early (in preschool children) and, then, can be modulated by factors linked to motor asymmetry and handedness. Finally, the study of manual lateralization in relation to language development has shown the importance of considering several characteristics of manual activities. In particular, the dissociation between manipulative activities and communicative gestures in young children may open further perspectives for future research on HS.

Quantifying the Confidence in fMRI-Based Language Lateralisation Through Laterality Index Deconstruction

In epilepsy patients, language lateralisation is an important part of the presurgical diagnostic process. Using task-based fMRI, language lateralisation can be determined by visual inspection of activity patterns or by quantifying the difference in left- and right-hemisphere activity using variations of a basic formula [(L-R)/(L+R)]. However, the values of this laterality index (LI) depend on the choice of activity thresholds and regions of interest. The diagnostic utility of the LI also depends on how its continuous values are translated into categorical decisions about a patient's language lateralisation. Here, we analysed fMRI data from 712 epilepsy patients who performed a verbal fluency task. Each fMRI data set was evaluated by a trained human rater as depicting left-sided, right-sided, or bilateral lateralisation or as being inconclusive. We used data-driven methods to define the activity thresholds and regions of interest used for LI computation and to define a classification scheme that allowed us to translate the LI values into categorical decisions. By deconstructing the LI into measures of laterality (L-R) and strength (L+R), we also modelled the relationship between activation strength and conclusiveness of a data set. In a held-out data set, predictions reached 91% correct when using only conclusive data and 82% when inconclusive data were included. Although only trained on human evaluations of fMRIs, the approach generalised to the prediction of language Wada test results, allowing for significant above-chance accuracies. Compared against different existing methods of LI-computation, our approach improved the identification and exclusion of inconclusive cases and ensured that decisions for the remaining data could be made with consistently high accuracies. We discuss how this approach can support clinicians in assessing fMRI data on a single-case level, deciding whether lateralisation can be determined with sufficient certainty or whether additional information is needed.

A SENtence Supramodal Areas AtlaS (SENSAAS) based on multiple task-induced activation mapping and graph analysis of intrinsic connectivity in 144 healthy right-handers

We herein propose an atlas of 32 sentence-related areas based on a 3-step method combining the analysis of activation and asymmetry during multiple language tasks with hierarchical clustering of resting-state connectivity and graph analyses. 144 healthy right-handers performed fMRI runs based on language production, reading and listening, both with sentences and lists of over-learned words. Sentence minus word-list BOLD contrast and left-minus-right BOLD asymmetry for each task were computed in pairs of homotopic regions of interest (hROIs) from the AICHA atlas. Thirty-two hROIs were identified that were conjointly activated and leftward asymmetrical in each of the three language contrasts. Analysis of resting-state temporal correlations of BOLD variations between these 32 hROIs allowed the segregation of a core network, SENT_CORE including 18 hROIs. Resting-state graph analysis applied to SENT_CORE hROIs revealed that the pars triangularis of the inferior frontal gyrus and the superior temporal sulcus were hubs based on their degree centrality (DC), betweenness, and participation values corresponding to epicentres of sentence processing. Positive correlations between DC and BOLD activation values for SENT_CORE hROIs were observed across individuals and across regions regardless of the task: the more a SENT_CORE area is connected at rest the stronger it is activated during sentence processing. DC measurements in SENT_CORE may thus be a valuable index for the evaluation of inter-individual variations in language areas functional activity in relation to anatomical or clinical patterns in large populations. SENSAAS (SENtence Supramodal Areas AtlaS), comprising the 32 supramodal sentence areas, including SENT_CORE network, can be downloaded at http://www.gin.cnrs.fr/en/tools/ .