Synaptic pruning in development: a computational account

Influencing factors of corticomuscular coherence in stroke patients

Stroke, also known as cerebrovascular accident, is an acute cerebrovascular disease with a high incidence, disability rate, and mortality. It can disrupt the interaction between the cerebral cortex and external muscles. Corticomuscular coherence (CMC) is a common and useful method for studying how the cerebral cortex controls muscle activity. CMC can expose functional connections between the cortex and muscle, reflecting the information flow in the motor system. Afferent feedback related to CMC can reveal these functional connections. This paper aims to investigate the factors influencing CMC in stroke patients and provide a comprehensive summary and analysis of the current research in this area. This paper begins by discussing the impact of stroke and the significance of CMC in stroke patients. It then proceeds to elaborate on the mechanism of CMC and its defining formula. Next, the impacts of various factors on CMC in stroke patients were discussed individually. Lastly, this paper addresses current challenges and future prospects for CMC.

A comparison of structural morphometry in children and adults with persistent developmental stuttering

This cross-sectional study aimed to differentiate earlier occurring neuroanatomical differences that may reflect core deficits in stuttering versus changes associated with a longer duration of stuttering by analysing structural morphometry in a large sample of children and adults who stutter and age-matched controls. Whole-brain T1-weighted structural scans were obtained from 166 individuals who stutter (74 children, 92 adults; ages 3-58) and 191 controls (92 children, 99 adults; ages 3-53) from eight prior studies in our laboratories. Mean size and gyrification measures were extracted using FreeSurfer software for each cortical region of interest. FreeSurfer software was also used to generate subcortical volumes for regions in the automatic subcortical segmentation. For cortical analyses, separate ANOVA analyses of size (surface area, cortical thickness) and gyrification (local gyrification index) measures were conducted to test for a main effect of diagnosis (stuttering, control) and the interaction of diagnosis-group with age-group (children, adults) across cortical regions. Cortical analyses were first conducted across a set of regions that comprise the speech network and then in a second whole-brain analysis. Next, separate ANOVA analyses of volume were conducted across subcortical regions in each hemisphere. False discovery rate corrections were applied for all analyses. Additionally, we tested for correlations between structural morphometry and stuttering severity. Analyses revealed thinner cortex in children who stutter compared with controls in several key speech-planning regions, with significant correlations between cortical thickness and stuttering severity. These differences in cortical size were not present in adults who stutter, who instead showed reduced gyrification in the right inferior frontal gyrus. Findings suggest that early cortical anomalies in key speech planning regions may be associated with stuttering onset. Persistent stuttering into adulthood may result from network-level dysfunction instead of focal differences in cortical morphometry. Adults who stutter may also have a more heterogeneous neural presentation than children who stutter due to their unique lived experiences.

Toward computational neuroconstructivism: a framework for developmental systems neuroscience

Brain development is underpinned by complex interactions between neural assemblies, driving structural and functional change. This neuroconstructivism (the notion that neural functions are shaped by these interactions) is core to some developmental theories. However, due to their complexity, understanding underlying developmental mechanisms is challenging. Elsewhere in neurobiology, a computational revolution has shown that mathematical models of hidden biological mechanisms can bridge observations with theory building. Can we build a similar computational framework yielding mechanistic insights for brain development? Here, we outline the conceptual and technical challenges of addressing this theory gap, and demonstrate that there is great potential in specifying brain development as mathematically defined processes operating within physical constraints. We provide examples, alongside broader ingredients needed, as the field explores computational explanations of system-wide development.

Microglial morphology in the somatosensory cortex across lifespan. A quantitative study

BACKGROUND

Microglia are long-lived cells that constantly monitor their microenvironment. To accomplish this task, they constantly change their morphology both in the short and long term under physiological conditions. This makes the process of quantifying physiological microglial morphology difficult.

RESULTS

By using a semi-manual and a semi-automatic method to assess fine changes in cortical microglia morphology, we were able to quantify microglia changes in number, surveillance and branch tree starting from the fifth postnatal day to 2 years of life. We were able to identify a fluctuating behavior of most analyzed parameters characterized by a rapid cellular maturation, followed by a long period of relative stable morphology during the adult life with a final convergence to an aged phenotype. Detailed cellular arborization analysis revealed age-induced differences in microglia morphology, with mean branch length and the number of terminal processes changing constantly over time.

CONCLUSIONS

Our study provides insight into microglia morphology changes across lifespan under physiological conditions. We were able to highlight, that due to the dynamic nature of microglia several morphological parameters are needed to establish the physiological state of these cells.

Neuroplasticity-Related Genes and Dopamine Receptors Associated with Regional Cortical Thickness Increase Following Electroconvulsive Therapy for Major Depressive Disorder

Electroconvulsive therapy (ECT) is an effective neuromodulatory therapy for major depressive disorder (MDD). Treatment is associated with regional changes in brain structure and function, indicating activation of neuroplastic processes. To investigate the underlying neurobiological mechanism of macroscopic reorganization following ECT, we longitudinally (before and after ECT in two centers) collected magnetic resonance images for 96 MDD patients. Similar patterns of cortical thickness (CT) changes following ECT were observed in two centers. These CT changes were spatially colocalized with a weighted combination of genes enriched for neuroplasticity-related ontology terms and pathways (e.g., synaptic pruning) as well as with a higher density of D2/3 dopamine receptors. A multiple linear regression model indicated that the region-specific gene expression and receptor density patterns explained 40% of the variance in CT changes after ECT. In conclusion, these findings suggested that dopamine signaling and neuroplasticity-related genes are associated with the ECT-induced morphological reorganization.

Pruning deep neural networks generates a sparse, bio-inspired nonlinear controller for insect flight

Insect flight is a strongly nonlinear and actuated dynamical system. As such, strategies for understanding its control have typically relied on either model-based methods or linearizations thereof. Here we develop a framework that combines model predictive control on an established flight dynamics model and deep neural networks (DNN) to create an efficient method for solving the inverse problem of flight control. We turn to natural systems for inspiration since they inherently demonstrate network pruning with the consequence of yielding more efficient networks for a specific set of tasks. This bio-inspired approach allows us to leverage network pruning to optimally sparsify a DNN architecture in order to perform flight tasks with as few neural connections as possible, however, there are limits to sparsification. Specifically, as the number of connections falls below a critical threshold, flight performance drops considerably. We develop sparsification paradigms and explore their limits for control tasks. Monte Carlo simulations also quantify the statistical distribution of network weights during pruning given initial random weights of the DNNs. We demonstrate that on average, the network can be pruned to retain a small amount of original network weights and still perform comparably to its fully-connected counterpart. The relative number of remaining weights, however, is highly dependent on the initial architecture and size of the network. Overall, this work shows that sparsely connected DNNs are capable of predicting the forces required to follow flight trajectories. Additionally, sparsification has sharp performance limits.

Cortical Thickness Abnormalities in Attention Deficit Hyperactivity Disorder Revealed by Structural Magnetic Resonance Imaging: Newborns to Young Adults

Attention deficit hyperactivity disorder is a neurodevelopmental condition for which we have an incomplete understanding, and so brain imaging methods, such as magnetic resonance imaging (MRI) may be able to assist in characterizing and understanding the presentation of the brain in an ADHD population. Statistical and computational methods were used to compare participants with attention deficit hyperactivity disorder (ADHD) and neurotypical controls at a variety of developmental stages to assess detectable abnormal neurodevelopment potentially associated with ADHD and to assess our ability to diagnose and characterize the condition from real-world clinical magnetic resonance imaging (MRI) examinations. T1-weighted structural MRI examinations (n=993; 0-31 years old [YO]) were obtained from neurotypical controls and 637 examinations were obtained from patients with ADHD (0-26 YO). Measures of average (mean) regional cortical thickness were acquired, alongside the first reporting of regional cortical thickness variability (as assessed with the standard deviation [SD]) in ADHD. A comparison between the inattentive and combined (inattentive and hyperactive) subtypes of ADHD is also provided. A preliminary independent validation was also performed on the publicly available ADHD200 dataset. Relative to controls, subjects with ADHD had, on average, lowered SD of cortical thicknesses and increased mean thicknesses across several key regions potentially linked with known symptoms of ADHD, including the precuneus, supramarginal gyrus, etc.

Exploring the development of high-level contributions to body representation using the rubber hand illusion and the monkey hand illusion

During development our body undergoes significant changes, yet we are able to maintain a coherent experience of our body and sense of self. Bodily experience is thought to comprise integration of multisensory signals (vision, touch, and proprioception) constrained by top-down knowledge of body appearance. Evidence from developmental studies suggests that low-level multisensory integration develops throughout childhood, reaching adult levels by 10 years of age. However, how high-level cognitive knowledge changes during childhood to constrain our multisensory body experience is unknown. This study describes four experiments examining high-level contributions to the bodily experience in children compared with adults using the rubber hand illusion and a monkey hand illusion. We found that children (5-17 years of age) exhibited more flexible body representations, showing stronger illusions for small and fantastical (monkey) fake hands compared with adults. Conversely, using a task indirectly capturing changes in hand size, we found that children and adults demonstrated statistically equivalent increases and decreases in hand size following illusions over large and small hands, respectively. Interestingly, at baseline children showed a bias in reporting larger hand size judgments that decreased with age. Finally, we did not find a relationship between individual differences in fantasy proneness and illusion strength for a fantastical (monkey) hand for children or adults, suggesting that developmental changes of top-down constraints are not purely driven by more diffuse boundaries between imagination and reality. These data suggest that high-level constraints acting on our multisensory body experience change during development, allowing children a more flexible bodily experience compared with adults.

A Synaptic Pruning-Based Spiking Neural Network for Hand-Written Digits Classification

A spiking neural network model inspired by synaptic pruning is developed and trained to extract features of hand-written digits. The network is composed of three spiking neural layers and one output neuron whose firing rate is used for classification. The model detects and collects the geometric features of the images from the Modified National Institute of Standards and Technology database (MNIST). In this work, a novel learning rule is developed to train the network to detect features of different digit classes. For this purpose, randomly initialized synaptic weights between the first and second layers are updated using average firing rates of pre- and postsynaptic neurons. Then, using a neuroscience-inspired mechanism named, “synaptic pruning” and its predefined threshold values, some of the synapses are deleted. Hence, these sparse matrices named, “information channels” are constructed so that they show highly specific patterns for each digit class as connection matrices between the first and second layers. The “information channels” are used in the test phase to assign a digit class to each test image. In addition, the role of feed-back inhibition as well as the connectivity rates of the second and third neural layers are studied. Similar to the abilities of the humans to learn from small training trials, the developed spiking neural network needs a very small dataset for training, compared to the conventional deep learning methods that have shown a very good performance on the MNIST dataset. This work introduces a new class of brain-inspired spiking neural networks to extract the features of complex data images.

Microglia sense neuronal activity via GABA in the early postnatal hippocampus

Microglia, the resident macrophages in the central nervous system, express receptors for classical neurotransmitters, such as γ-aminobutyric acid (GABA) and glutamate, suggesting that they sense synaptic activity. To detect microglial Ca²⁺ responses to neuronal activity, we generate transgenic mouse lines expressing the fluorescent Ca²⁺ indicator GCaMP6m, specifically in microglia and demonstrate that electrical stimulation of the Schaffer collateral pathway results in microglial Ca²⁺ responses in early postnatal but not adult hippocampus. Preceding the microglial responses, we also observe similar Ca²⁺ responses in astrocytes, and both are sensitive to tetrodotoxin. Blocking astrocytic glutamate uptake or GABA transport abolishes stimulation-induced microglial responses as well as antagonizing the microglial GABA_B receptor. Our data, therefore, suggest that the neuronal activity-induced glutamate uptake and the release of GABA by astrocytes trigger the activation of GABA_B receptors in microglia. This neuron, astrocyte, and microglia communication pathway might modulate microglial activity in developing neuronal networks.

Computational benefits of structural plasticity, illustrated in songbirds

The plasticity of nervous systems allows animals to quickly adapt to a changing environment. In particular, the structural plasticity of brain networks is often critical to the development of the central nervous system and the acquisition of complex behaviors. As an example, structural plasticity is central to the development of song-related brain circuits and may be critical for song acquisition in juvenile songbirds. Here, we review current evidences for structural plasticity and their significance from a computational point of view. We start by reviewing evidence for structural plasticity across species and categorizing them along the spatial axes as well as the along the time course during development. We introduce the vocal learning circuitry in zebra finches, as a useful example of structural plasticity, and use this specific case to explore the possible contributions of structural plasticity to computational models. Finally, we discuss current modeling studies incorporating structural plasticity and unexplored questions which are raised by such models.

Role of NMDAR plasticity in a computational model of synaptic memory

A largely unexplored question in neuronal plasticity is whether synapses are capable of encoding and learning the timing of synaptic inputs. We address this question in a computational model of synaptic input time difference learning (SITDL), where N-methyl-d-aspartate receptor (NMDAR) isoform expression in silent synapses is affected by time differences between glutamate and voltage signals. We suggest that differences between NMDARs' glutamate and voltage gate conductances induce modifications of the synapse's NMDAR isoform population, consequently changing the timing of synaptic response. NMDAR expression at individual synapses can encode the precise time difference between signals. Thus, SITDL enables the learning and reconstruction of signals across multiple synapses of a single neuron. In addition to plausibly predicting the roles of NMDARs in synaptic plasticity, SITDL can be usefully applied in artificial neural network models.

Toward a Brain-Inspired Developmental Neural Network Based on Dendritic Spine Dynamics

Neural networks with a large number of parameters are prone to overfitting problems when trained on a relatively small training set. Introducing weight penalties of regularization is a promising technique for solving this problem. Taking inspiration from the dynamic plasticity of dendritic spines, which plays an important role in the maintenance of memory, this letter proposes a brain-inspired developmental neural network based on dendritic spine dynamics (BDNN-dsd). The dynamic structure changes of dendritic spines include appearing, enlarging, shrinking, and disappearing. Such spine plasticity depends on synaptic activity and can be modulated by experiences-in particular, long-lasting synaptic enhancement/suppression (LTP/LTD), coupled with synapse formation (or enlargement)/elimination (or shrinkage), respectively. Subsequently, spine density characterizes an approximate estimate of the total number of synapses between neurons. Motivated by this, we constrain the weight to a tunable bound that can be adaptively modulated based on synaptic activity. Dynamic weight bound could limit the relatively redundant synapses and facilitate the contributing synapses. Extensive experiments demonstrate the effectiveness of our method on classification tasks of different complexity with the MNIST, Fashion MNIST, and CIFAR-10 data sets. Furthermore, compared to dropout and L2 regularization algorithms, our method can improve the network convergence rate and classification performance even for a compact network.

The information theory of developmental pruning: Optimizing global network architectures using local synaptic rules

During development, biological neural networks produce more synapses and neurons than needed. Many of these synapses and neurons are later removed in a process known as neural pruning. Why networks should initially be over-populated, and the processes that determine which synapses and neurons are ultimately pruned, remains unclear. We study the mechanisms and significance of neural pruning in model neural networks. In a deep Boltzmann machine model of sensory encoding, we find that (1) synaptic pruning is necessary to learn efficient network architectures that retain computationally-relevant connections, (2) pruning by synaptic weight alone does not optimize network size and (3) pruning based on a locally-available measure of importance based on Fisher information allows the network to identify structurally important vs. unimportant connections and neurons. This locally-available measure of importance has a biological interpretation in terms of the correlations between presynaptic and postsynaptic neurons, and implies an efficient activity-driven pruning rule. Overall, we show how local activity-dependent synaptic pruning can solve the global problem of optimizing a network architecture. We relate these findings to biology as follows: (I) Synaptic over-production is necessary for activity-dependent connectivity optimization. (II) In networks that have more neurons than needed, cells compete for activity, and only the most important and selective neurons are retained. (III) Cells may also be pruned due to a loss of synapses on their axons. This occurs when the information they convey is not relevant to the target population.

Biological neural networks need to be efficient and compact, as synapses and neurons require space to store and energy to operate and maintain. This favors an optimized network topology that minimizes redundant neurons and connections. Large numbers of extra neurons and synapses are produced during development, and later removed as the brain matures. A key question to understand this process is how neurons determine which synapses are important. We used statistical models of neural networks to simulate developmental pruning. We show that neurons in such networks can use locally available information to measure the importance of their synapses in a biologically plausible way. We demonstrate that this pruning rule, which is motivated by information theoretic considerations, retains network topologies that can efficiently encode sensory inputs. In contrast, pruning at random, or based on synaptic weights alone, was less able to identify redundant neurons.

Dynamically Optimizing Network Structure Based on Synaptic Pruning in the Brain

Most neural networks need to predefine the network architecture empirically, which may cause over-fitting or under-fitting. Besides, a large number of parameters in a fully connected network leads to the prohibitively expensive computational cost and storage overhead, which makes the model hard to be deployed on mobile devices. Dynamically optimizing the network architecture by pruning unused synapses is a promising technique for solving this problem. Most existing pruning methods focus on reducing the redundancy of deep convolutional neural networks by pruning unimportant filters or weights, at the cost of accuracy drop. In this paper, we propose an effective brain-inspired synaptic pruning method to dynamically modulate the network architecture and simultaneously improve network performance. The proposed model is biologically inspired as it dynamically eliminates redundant connections based on the synaptic pruning rules used during the brain's development. Connections are pruned if they are not activated or less activated multiple times consecutively. Extensive experiments demonstrate the effectiveness of our method on classification tasks of different complexity with the MNIST, Fashion MNIST, and CIFAR-10 datasets. Experimental results reveal that even for a compact network, the proposed method can also remove up to 59-90% of the connections, with relative improvement in learning speed and accuracy.

Environmental influences on the pace of brain development

Childhood socio-economic status (SES), a measure of the availability of material and social resources, is one of the strongest predictors of lifelong well-being. Here we review evidence that experiences associated with childhood SES affect not only the outcome but also the pace of brain development. We argue that higher childhood SES is associated with protracted structural brain development and a prolonged trajectory of functional network segregation, ultimately leading to more efficient cortical networks in adulthood. We hypothesize that greater exposure to chronic stress accelerates brain maturation, whereas greater access to novel positive experiences decelerates maturation. We discuss the impact of variation in the pace of brain development on plasticity and learning. We provide a generative theoretical framework to catalyse future basic science and translational research on environmental influences on brain development.

Growth strategy determines the memory and structural properties of brain networks

The interplay between structure and function affects the emerging properties of many natural systems. Here we use an adaptive neural network model that couples activity and topological dynamics and reproduces the experimental temporal profiles of synaptic density observed in the brain. We prove that the existence of a transient period of relatively high synaptic connectivity is critical for the development of the system under noise circumstances, such that the resulting network can recover stored memories. Moreover, we show that intermediate synaptic densities provide optimal developmental paths with minimum energy consumption, and that ultimately it is the transient heterogeneity in the network that determines its evolution. These results could explain why the pruning curves observed in actual brain areas present their characteristic temporal profiles and they also suggest new design strategies to build biologically inspired neural networks with particular information processing capabilities.

Developmental Changes in ANS Precision Across Grades 1-9: Different Patterns of Accuracy and Reaction Time

The main aim of this study was to analyze the patterns of changes in Approximate Number Sense (ANS) precision from grade 1 (mean age: 7.84 years) to grade 9 (mean age: 15.82 years) in a sample of Russian schoolchildren. To fulfill this aim, the data from a longitudinal study of two cohorts of children were used. The first cohort was assessed at grades 1-5 (elementary school education plus the first year of secondary education), and the second cohort was assessed at grades 5-9 (secondary school education). ANS precision was assessed by accuracy and reaction time (RT) in a non-symbolic comparison test ("blue-yellow dots" test). The patterns of change were estimated via mixed-effect growth models. The results revealed that in the first cohort, the average accuracy increased from grade 1 to grade 5 following a non-linear pattern and that the rate of growth slowed after grade 3 (7-9 years old). The non-linear pattern of changes in the second cohort indicated that accuracy started to increase from grade 7 to grade 9 (13-15 years old), while there were no changes from grade 5 to grade 7. However, the RT in the non-symbolic comparison test decreased evenly from grade 1 to grade 7 (7-13 years old), and the rate of processing non-symbolic information tended to stabilize from grade 7 to grade 9. Moreover, the changes in the rate of processing non-symbolic information were not explained by the changes in general processing speed. The results also demonstrated that accuracy and RT were positively correlated across all grades. These results indicate that accuracy and the rate of non-symbolic processing reflect two different processes, namely, the maturation and development of a non-symbolic representation system.

Visual cortical plasticity and the risk for psychosis: An interim analysis of the North American Prodrome Longitudinal Study

BACKGROUND

Adolescence/early adulthood coincides with accelerated pruning of cortical synapses and the onset of schizophrenia. Cortical gray matter reduction and dysconnectivity in schizophrenia are hypothesized to result from impaired synaptic plasticity mechanisms, including long-term potentiation (LTP), since deficient LTP may result in too many weak synapses that are then subject to over-pruning. Deficient plasticity has already been observed in schizophrenia. Here, we assessed whether such deficits are present in the psychosis risk syndrome (PRS), particularly those who subsequently convert to full psychosis.

METHODS

An interim analysis was performed on a sub-sample from the NAPLS-3 study, including 46 healthy controls (HC) and 246 PRS participants. All participants performed an LTP-like visual cortical plasticity paradigm involving assessment of visual evoked potentials (VEPs) elicited by vertical and horizontal line gratings before and after high frequency ("tetanizing") visual stimulation with one of the gratings to induce "input-specific" neuroplasticity (i.e., VEP changes specific to the tetanized stimulus). Non-parametric, cluster-based permutation testing was used to identify electrodes and timepoints that demonstrated input-specific plasticity effects.

RESULTS

Input-specific pre-post VEP changes (i.e., increased negative voltage) were found in a single spatio-temporal cluster covering multiple occipital electrodes in a 126-223 ms time window. This plasticity effect was deficient in PRS individuals who subsequently converted to psychosis, relative to PRS non-converters and HC.

CONCLUSIONS

Input-specific LTP-like visual plasticity can be measured from VEPs in adolescents and young adults. Interim analyses suggest that deficient visual cortical plasticity is evident in those PRS individuals at greatest risk for transition to psychosis.

A Comparative Biology of Microglia Across Species

Microglia are unique brain-resident, myeloid cells. They have received growing interest for their implication in an increasing number of neurodevelopmental, acute injury, and neurodegenerative disorders of the central nervous system (CNS). Fate-mapping studies establish microglial ontogeny from the periphery during development, while recent transcriptomic studies highlight microglial identity as distinct from other CNS cells and peripheral myeloid cells. This evidence for a unique microglial ontogeny and identity raises questions regarding their identity and functions across species. This review will examine the available evidence for microglia in invertebrate and vertebrate species to clarify similarities and differences in microglial identity, ontogeny, and physiology across species. This discussion highlights conserved and divergent microglial properties through evolution. Finally, we suggest several interesting research directions from an evolutionary perspective to adequately understand the significance of microglia emergence. A proper appreciation of microglia from this perspective could inform the development of specific therapies geared at targeting microglia in various pathologies.

Maternal depression trajectories and child executive function over 9 years

BACKGROUND

The present study aimed to assess the relationship between maternal depression trajectories from pregnancy to 2 years after childbirth and childhood behavioral problems and executive function at 9 years.

METHODS

Data of mother-child pairs (N = 1191) extracted from the Panel Study on Korean Children (a cohort study) were used. Maternal depression was assessed using the Kessler depression scale during pregnancy and at 6 months, 1 years, and 2 years postpartum. At ages 4, 5, 6, 7, and 9 years, the children's behavioral outcomes were assessed using the Child Behavior Checklist. The children's executive function was assessed using the Executive Function Difficulty Screening Questionnaire at ages 7, 8, and 9 years. We performed a latent profile analysis to identify maternal depression trajectories and compared the children's behavioral problems and executive function among different trajectories.

RESULTS

According to maternal depression trajectory, the mother-child pairs were divided into the no symptom (n = 503), mild symptom (n = 558), and moderate symptom (n = 130) groups. Children of mothers with significant depressive symptoms had severe behavioral problems at ages 4, 5, 6, 7, and 9 years. Moreover, compared with children whose mothers were not depressed, those whose mothers had mild or moderate symptoms had impaired executive function at ages 7, 8, and 9 years.

CONCLUSIONS

Maternal depression up to 2 years after childbirth affects childhood behavior and executive function into middle childhood.

One-week phonemic training rebuilds the memory traces of merged phonemes in merged speakers

The phonemic merger is a unique phenomenon which is referred to as acoustically very different phonemes are recognized as the same phoneme. In our previous study, we demonstrated that the merged speakers had lost the ability to discriminate the merged phonemes pre-attentively, as revealed by their failure in mismatch negativity (MMN) elicitation in the oddball stream of the merged phonemes /n/-/l/. In this study, we investigated the recovery of the discrimination ability via phonemic training and found that the merged speakers regained the ability of discriminating merged phonemes pre-attentively, after a 7-day /n/-/l/ phonemic training, as revealed by the reactivation of MMN brain response to the /n/-/l/ phoneme categories. Our finding indicates that separate memory traces of merged phonemes could be rebuilt during the training process.

The effect of bilingualism on brain development from early childhood to young adulthood

Bilingualism affects the structure of the brain in adults, as evidenced by experience-dependent grey and white matter changes in brain structures implicated in language learning, processing, and control. However, limited evidence exists on how bilingualism may influence brain development. We examined the developmental patterns of both grey and white matter structures in a cross-sectional study of a large sample (n = 711 for grey matter, n = 637 for white matter) of bilingual and monolingual participants, aged 3–21 years. Metrics of grey matter (thickness, volume, and surface area) and white matter (fractional anisotropy and mean diffusivity) were examined across 41 cortical and subcortical brain structures and 20 tracts, respectively. We used generalized additive modelling to analyze whether, how, and where the developmental trajectories of bilinguals and monolinguals might differ. Bilingual and monolingual participants manifested distinct developmental trajectories in both grey and white matter structures. As compared to monolinguals, bilinguals showed: (a) more grey matter (less developmental loss) starting during late childhood and adolescence, mainly in frontal and parietal regions (particularly in the inferior frontal gyrus pars opercularis, superior frontal cortex, inferior and superior parietal cortex, and precuneus); and (b) higher white matter integrity (greater developmental increase) starting during mid-late adolescence, specifically in striatal–inferior frontal fibers. The data suggest that there may be a developmental basis to the well-documented structural differences in the brain between bilingual and monolingual adults.

Altered Corticomuscular Coherence (CMCoh) Pattern in the Upper Limb During Finger Movements After Stroke

Background: Proximal compensation to the distal movements is commonly observed in the affected upper extremity (UE) of patients with chronic stroke. However, the cortical origin of this compensation has not been well-understood. In this study, corticomuscular coherence (CMCoh) and electromyography (EMG) analysis were adopted to investigate the corticomuscular coordinating pattern of proximal UE compensatory activities when conducting distal UE movements in chronic stroke.

Method: Fourteen chronic stroke subjects and 10 age-matched unimpaired controls conducted isometric finger extensions and flexions at 20 and 40% of maximal voluntary contractions. Electroencephalogram (EEG) data were recorded from the sensorimotor area and EMG signals were captured from extensor digitorum (ED), flexor digitorum (FD), triceps brachii (TRI), and biceps brachii (BIC) to investigate the CMCoh peak values in the Beta band. EMG parameters, i.e., the EMG activation level and co-contraction index (CI), were analyzed to evaluate the compensatory muscular patterns in the upper limb.

Result: The peak CMCoh with statistical significance (P < 0.05) was found shifted from the ipsilesional side to the contralesional side in the proximal UE muscles, while to the central regions in the distal UE muscle in chronic strokes. Significant differences (P < 0.05) were observed in both peak ED and FD CMCohs during finger extensions between the two groups. The unimpaired controls exhibited significant intragroup differences between 20 and 40% levels in extensions for peak ED and FD CMCohs (P < 0.05). The stroke subjects showed significant differences in peak TRI and BIC CMCohs (P < 0.01). No significant inter- or intra-group difference was observed in peak CMCoh during finger flexions. EMG parameters showed higher EMG activation levels in TRI and BIC muscles (P < 0.05), and higher CI values in the muscle pairs involving TRI and BIC during all the extension and flexion tasks in the stroke group than those in the control group (P < 0.05).

Conclusion: The post-stroke proximal muscular compensations from the elbow to the finger movements were cortically originated, with the center mainly located in the contralesional hemisphere.

ThiNet: Pruning CNN Filters for a Thinner Net

This paper aims at accelerating and compressing deep neural networks to deploy CNN models into small devices like mobile phones or embedded gadgets. We focus on filter level pruning, i.e., the whole filter will be discarded if it is less important. An effective and unified framework, ThiNet (stands for "Thin Net"), is proposed in this paper. We formally establish filter pruning as an optimization problem, and reveal that we need to prune filters based on statistics computed from its next layer, not the current layer, which differentiates ThiNet from existing methods. We also propose "gcos" (Group COnvolution with Shuffling), a more accurate group convolution scheme, to further reduce the pruned model size. Experimental results demonstrate the effectiveness of our method, which has advanced the state-of-the-art. Moreover, we show that the original VGG-16 model can be compressed into a very small model (ThiNet-Tiny) with only 2.66 MB model size, but still preserve AlexNet level accuracy. This small model is evaluated on several benchmarks with different vision tasks (e.g., classification, detection, segmentation), and shows excellent generalization ability.

How Memory Conforms to Brain Development

Nature exhibits countless examples of adaptive networks, whose topology evolves constantly coupled with the activity due to its function. The brain is an illustrative example of a system in which a dynamic complex network develops by the generation and pruning of synaptic contacts between neurons while memories are acquired and consolidated. Here, we consider a recently proposed brain developing model to study how mechanisms responsible for the evolution of brain structure affect and are affected by memory storage processes. Following recent experimental observations, we assume that the basic rules for adding and removing synapses depend on local synaptic currents at the respective neurons in addition to global mechanisms depending on the mean connectivity. In this way a feedback loop between "form" and "function" spontaneously emerges that influences the ability of the system to optimally store and retrieve sensory information in patterns of brain activity or memories. In particular, we report here that, as a consequence of such a feedback-loop, oscillations in the activity of the system among the memorized patterns can occur, depending on parameters, reminding mind dynamical processes. Such oscillations have their origin in the destabilization of memory attractors due to the pruning dynamics, which induces a kind of structural disorder or noise in the system at a long-term scale. This constantly modifies the synaptic disorder induced by the interference among the many patterns of activity memorized in the system. Such new intriguing oscillatory behavior is to be associated only to long-term synaptic mechanisms during the network evolution dynamics, and it does not depend on short-term synaptic processes, as assumed in other studies, that are not present in our model.

Emotional Theory of Rationality

In recent decades, the existence of a close relationship between emotional phenomena and rational processes has certainly been established, yet there is still no unified definition or effective model to describe them. To advance our understanding of the mechanisms governing the behavior of living beings, we must integrate multiple theories, experiments, and models from both fields. In this article we propose a new theoretical framework that allows integrating and understanding the emotion-cognition duality, from a functional point of view. Based on evolutionary principles, our reasoning adds to the definition and understanding of emotion, justifying its origin, explaining its mission and dynamics, and linking it to higher cognitive processes, mainly with attention, cognition, decision-making, and consciousness. According to our theory, emotions are the mechanism for brain function optimization, aside from the contingency and stimuli prioritization system. As a result of this approach, we have developed a dynamic systems-level model capable of providing plausible explanations for certain psychological and behavioral phenomena and establishing a new framework for the scientific definition of some fundamental psychological terms.

Pubertal hormones increase hippocampal expression of α4βδ GABAA receptors

CA1 hippocampal expression of α4βδ GABA_A receptors (GABARs) increases at the onset of puberty in female mice, an effect dependent upon the decline in hippocampal levels of the neurosteroid THP (3α-OH-5α-pregnan-20-one) which occurs at this time. The present study further characterized the mechanisms underlying α4βδ expression, assessed in vivo. Blockade of pubertal levels of 17β-estradiol (E₂) (formestane, 0.5 mg/kg, i.p. 3 d) reduced α4 and δ expression by 75-80% (P < 0.05) in CA1 hippocampus of female mice, assessed using Western blot techniques. Conversely, E₂ administration increased α4 and δ expression by 50-100% in adults, an effect enhanced by more than 2-fold by concomitant administration of the 5α-reductase blocker finasteride (50 mg/kg, i.p., 3d, P < 0.05), suggesting that both declining THP levels and increasing E₂ levels before puberty trigger α4βδ expression. This effect was blocked by ICI 182,780 (20 mg/kg, s.c., 3 d), a selective blocker of E₂ receptor-α (ER-α). These results suggest that both the rise in circulating levels of E₂ and the decline in hippocampal THP levels at the onset of puberty trigger maximal levels of α4βδ expression in the CA1 hippocampus.

α4βδ GABAA Receptors Trigger Synaptic Pruning and Reduce Dendritic Length of Female Mouse CA3 Hippocampal Pyramidal Cells at Puberty

Synaptic pruning during adolescence is critical for optimal cognition. The CA3 hippocampus contains unique spine types and plays a pivotal role in pattern separation and seizure generation, where sex differences exist, but adolescent pruning has only been studied in the male. Thus, for the present study we assessed pruning of specific spine types in the CA3 hippocampus during adolescence and investigated a possible mechanism in the female mouse. To this end, we used Golgi-impregnated brains from pubertal (∼PND 35, assessed by vaginal opening) and post-pubertal (PND 56) mice. Spine density was assessed from z-stack (0.1-μm steps) images taken using a Nikon DS-U3 camera through a Nikon Eclipse Ci-L microscope and analyzed with NIS Elements. Spine density decreased significantly (P < 0.05) during adolescence, with 50-60% decreases in mushroom and stubby spine-types (P < 0.05, ∼PND35 vs. PND56) in non-proestrous mice. This was associated with decreases in kalirin-7, a spine protein which stabilizes the cytoskeleton and is required for spine maintenance. Because our previous findings suggest that pubertal increases in α4βδ GABAA receptors (GABARs) trigger pruning in CA1, we investigated their role in CA3. α4 expression in CA3 hippocampus increased 4-fold at puberty (P < 0.05), assessed by immunostaining and verified electrophysiologically by an increased response to gaboxadol (100 nM), which is selective for α4βδ. Knock-out of α4 prevented the pubertal decrease in kalirin-7 and synaptic pruning and also increased the dendritic length, demonstrating a functional link. These data suggest that pubertal α4βδ GABARs alter dendritic morphology and trigger pruning in female CA3 hippocampus.

Prefrontal and Hippocampal Structure Predict Statistical Learning Ability in Early Childhood

Statistical learning can be used to gain sensitivity to many important regularities in our environment, including structure that is foundational to language and visual perception. As yet, little is known about how statistical learning takes place in the human brain, especially in children's developing brains and with regard to the broader neurobiology of learning and memory. We therefore explored the relationship between statistical learning and the thickness and volume of structures that are traditionally implicated in declarative and procedural memory, focusing specifically on the left inferior PFC, the hippocampus, and the caudate during early childhood (ages 5-8.5 years). We found that the thickness of the left inferior frontal cortex and volume of the right hippocampus predicted statistical learning ability in young children. Importantly, these regions did not change in thickness or volume with age, but the relationship between learning and the right hippocampus interacted with age such that older children's hippocampal structure more strongly predicted performance. Overall, the data show that children's statistical learning is supported by multiple neural structures that are more broadly implicated in learning and memory, especially declarative memory (hippocampus) and attention/top-down control (the PFC).

Recurrent Spiking Neural Network Learning Based on a Competitive Maximization of Neuronal Activity

Spiking neural networks (SNNs) are believed to be highly computationally and energy efficient for specific neurochip hardware real-time solutions. However, there is a lack of learning algorithms for complex SNNs with recurrent connections, comparable in efficiency with back-propagation techniques and capable of unsupervised training. Here we suppose that each neuron in a biological neural network tends to maximize its activity in competition with other neurons, and put this principle at the basis of a new SNN learning algorithm. In such a way, a spiking network with the learned feed-forward, reciprocal and intralayer inhibitory connections, is introduced to the MNIST database digit recognition. It has been demonstrated that this SNN can be trained without a teacher, after a short supervised initialization of weights by the same algorithm. Also, it has been shown that neurons are grouped into families of hierarchical structures, corresponding to different digit classes and their associations. This property is expected to be useful to reduce the number of layers in deep neural networks and modeling the formation of various functional structures in a biological nervous system. Comparison of the learning properties of the suggested algorithm, with those of the Sparse Distributed Representation approach shows similarity in coding but also some advantages of the former. The basic principle of the proposed algorithm is believed to be practically applicable to the construction of much more complicated and diverse task solving SNNs. We refer to this new approach as "Family-Engaged Execution and Learning of Induced Neuron Groups," or FEELING.

Rett Syndrome: Treatment with IGF-I, Melatonin, Blackcurrant Extracts, and Rehabilitation

(1) This study describes the good evolution of a 6-year-old girl genetically diagnosed (R106X) with Rett syndrome (RTT), after having been treated with IGF-I, melatonin (MT), blackcurrant extracts (BC) and rehabilitated for 6 months. (2) The patient stopped normal development in the first year of age. The patient showed short stature and weight and fulfilled the main criteria for typical RTT. Despite her young age, there was pubic hair (Tanner II), very high plasma testosterone, and low levels of plasma gonadotrophins. There were no adrenal enzymatic deficits, and abdominal ultrasound studies were normal. The treatment consisted of IGF-I (0.04 mg/kg/day, 5 days/week, subcutaneous (sc)) for 3 months and then 15 days of rest, MT (50 mg/day, orally, without interruption) and neurorehabilitation. A new blood test, after 3 months of treatment, was absolutely normal and the pubic hair disappeared (Tanner I). Then, a new treatment was started with IGF-I, MT, and BC for another 3 months. In this period, the degree of pubertal development increased to Tanner III (pubic level), without a known cause. (3) The treatment followed led to clear improvements in most of the initial abnormalities, perhaps due to the neurotrophic effect of IGF-I, the antioxidant effects of MT and BC, and the cerebral increase in the cyclic glycine-proline (cGP) achieved with administration of BC. (4) A continuous treatment with IGF-I, MT, and BC appears to be useful in RTT.

Do patterns of synaptic pruning underlie psychoses, autism and ADHD?

This review of available longitudinal structural imaging and immunological findings in first-onset schizophrenia, bipolar disorder, attention-deficit hyperactivity disorder (ADHD) and autism suggests that different patterns of synaptic pruning lead to various phenotypes. Proposals for future research strategies to try to replicate these findings are suggested, potential biomarkers to assist in diagnosis and determining the optimum duration of maintenance treatment are considered and ideas of potential immunotherapy augmentation are outlined.

LEARNING OBJECTIVES

•

DECLARATION OF INTEREST

None.

Development of Brain Network in Children with Autism from Early Childhood to Late Childhood

Extensive studies have indicated brain function connectivity abnormalities in autism spectrum disorder (ASD). However, there is a lack of longitudinal or cross-sectional research focused on tracking age-related developmental trends of autistic children at an early stage of brain development or based on a relatively large sample. The present study examined brain network changes in a total of 186 children both with and without ASD from 3 to 11 years, an early and key development period when significant changes are expected. The study aimed to investigate possible abnormal connectivity patterns and topological properties of children with ASD from early childhood to late childhood by using resting-state electroencephalographic (EEG) data. The main findings of the study were as follows: (1) From the connectivity analysis, several inter-regional synchronizations with reduction were identified in the younger and older ASD groups, and several intra-regional synchronization increases were observed in the older ASD group. (2) From the graph analysis, a reduced clustering coefficient and enhanced mean shortest path length in specific frequencies was observed in children with ASD. (3) Results suggested an age-related decrease of the mean shortest path length in the delta and theta bands in TD children, whereas atypical age-related alteration was observed in the ASD group. In addition, graph measures were correlated with ASD symptom severity in the alpha band. These results demonstrate that abnormal neural communication is already present at the early stages of brain development in autistic children and this may be involved in the behavioral deficits associated with ASD.

Network Design and the Brain

Neural circuits have evolved to accommodate similar information processing challenges as those faced by engineered systems. Here, we compare neural versus engineering strategies for constructing networks. During circuit development, synapses are overproduced and then pruned back over time, whereas in engineered networks, connections are initially sparse and are then added over time. We provide a computational perspective on these two different approaches, including discussion of how and why they are used, insights that one can provide the other, and areas for future joint investigation. By thinking algorithmically about the goals, constraints, and optimization principles used by neural circuits, we can develop brain-derived strategies for enhancing network design, while also stimulating experimental hypotheses about circuit development and function.

Limited synapse overproduction can speed development but sometimes with long-term energy and discrimination penalties

Neural circuit development requires that synapses be formed between appropriate neurons. In addition, for a hierarchical network, successful development involves a sequencing of developmental events. It has been suggested that one mechanism that helps speed up development of proper connections is an early overproduction of synapses. Using a computational model of synapse development, such as adaptive synaptogenesis, it is possible to study such overproduction and its role in speeding up development; it is also possible to study other outcomes of synapse overproduction that are seemingly new to the literature. With a fixed number of neurons, adaptive synaptogenesis can control the speed of synaptic development in two ways: by altering the rate constants of the adaptive processes or by altering the initial number of rapidly but non-selectively accrued synapses. Using either mechanism, the simulations reveal that synapse overproduction appears as an unavoidable concomitant of rapid adaptive synaptogenesis. However, the shortest development times, which always produces the greatest amount of synapse overproduction, reduce adult performance by three measures: energy use, discrimination error rates, and proportional neuron allocation. Thus, the results here lead to the hypothesis that the observed speed of neural network development represents a particular inter-generational compromise: quick development benefits parental fecundity while slow development benefits offspring fecundity.

Altered right anterior insular connectivity and loss of associated functions in adolescent chronic fatigue syndrome

Impairments in cognition, pain intolerance, and physical inactivity characterize adolescent chronic fatigue syndrome (CFS), yet little is known about its neurobiology. The right dorsal anterior insular (dAI) connectivity of the salience network provides a motivational context to stimuli. In this study, we examined regional functional connectivity (FC) patterns of the right dAI in adolescent CFS patients and healthy participants. Eighteen adolescent patients with CFS and 18 aged-matched healthy adolescent control participants underwent resting-state functional magnetic resonance imaging. The right dAI region of interest was examined in a seed-to-voxel resting-state FC analysis using SPM and CONN toolbox. Relative to healthy adolescents, CFS patients demonstrated reduced FC of the right dAI to the right posterior parietal cortex (PPC) node of the central executive network. The decreased FC of the right dAI–PPC might indicate impaired cognitive control development in adolescent CFS. Immature FC of the right dAI–PPC in patients also lacked associations with three known functional domains: cognition, pain and physical activity, which were observed in the healthy group. These results suggest a distinct biological signature of adolescent CFS and might represent a fundamental role of the dAI in motivated behavior.

Microglial Function across the Spectrum of Age and Gender

Microglia constitute the resident immunocompetent cells of the central nervous system. Although much work has focused on their ability to mount an inflammatory response in reaction to pathology, recent studies have delved into their role in maintaining homeostasis in the healthy brain. It is important to note that the function of these cells is more complex than originally conceived, as there is increasing evidence that microglial responses can vary greatly among individuals. Here, this review will describe the changing behavior of microglia from development and birth through to the aged brain. Further, it is not only age that impacts the state of the neuroimmune milieu, as microglia have been shown to play a central role in the sexual differentiation of the brain. Finally, this review will discuss the implications this has for the differences in the incidence of neurodegenerative disorders between males and females, and between the young and old.

Development of neural population activity toward self-organized criticality

Self-organized criticality (SoC), a spontaneous dynamic state established and maintained in networks of moderate complexity, is a universal characteristic of neural systems. Such systems produce cascades of spontaneous activity that are typically characterized by power-law distributions and rich, stable spatiotemporal patterns (i.e., neuronal avalanches). Since the dynamics of the critical state confer advantages in information processing within neuronal networks, it is of great interest to determine how criticality emerges during development. One possible mechanism is developmental, and includes axonal elongation during synaptogenesis and subsequent synaptic pruning in combination with the maturation of GABAergic inhibition (i.e., the integration then fragmentation process). Because experimental evidence for this mechanism remains inconclusive, we studied the developmental variation of neuronal avalanches in dissociated cortical neurons using high-density complementary metal-oxide semiconductor (CMOS) microelectrode arrays (MEAs). The spontaneous activities of nine cultures were monitored using CMOS MEAs from 4 to 30days in vitro (DIV) at single-cell spatial resolution. While cells were immature, cultures demonstrated random-like patterns of activity and an exponential avalanche size distribution; this distribution was followed by a bimodal distribution, and finally a power-law-like distribution. The bimodal distribution was associated with a large-scale avalanche with a homogeneous spatiotemporal pattern, while the subsequent power-law distribution was associated with diverse patterns. These results suggest that the SoC emerges through a two-step process: the integration process accompanying the characteristic large-scale avalanche and the fragmentation process associated with diverse middle-size avalanches.

The Indispensable Roles of Microglia and Astrocytes during Brain Development

Glia are essential for brain functioning during development and in the adult brain. Here, we discuss the various roles of both microglia and astrocytes, and their interactions during brain development. Although both cells are fundamentally different in origin and function, they often affect the same developmental processes such as neuro-/gliogenesis, angiogenesis, axonal outgrowth, synaptogenesis and synaptic pruning. Due to their important instructive roles in these processes, dysfunction of microglia or astrocytes during brain development could contribute to neurodevelopmental disorders and potentially even late-onset neuropathology. A better understanding of the origin, differentiation process and developmental functions of microglia and astrocytes will help to fully appreciate their role both in the developing as well as in the adult brain, in health and disease.

Structural Plasticity Denoises Responses and Improves Learning Speed

Despite an abundance of computational models for learning of synaptic weights, there has been relatively little research on structural plasticity, i.e., the creation and elimination of synapses. Especially, it is not clear how structural plasticity works in concert with spike-timing-dependent plasticity (STDP) and what advantages their combination offers. Here we present a fairly large-scale functional model that uses leaky integrate-and-fire neurons, STDP, homeostasis, recurrent connections, and structural plasticity to learn the input encoding, the relation between inputs, and to infer missing inputs. Using this model, we compare the error and the amount of noise in the network's responses with and without structural plasticity and the influence of structural plasticity on the learning speed of the network. Using structural plasticity during learning shows good results for learning the representation of input values, i.e., structural plasticity strongly reduces the noise of the response by preventing spikes with a high error. For inferring missing inputs we see similar results, with responses having less noise if the network was trained using structural plasticity. Additionally, using structural plasticity with pruning significantly decreased the time to learn weights suitable for inference. Presumably, this is due to the clearer signal containing less spikes that misrepresent the desired value. Therefore, this work shows that structural plasticity is not only able to improve upon the performance using STDP without structural plasticity but also speeds up learning. Additionally, it addresses the practical problem of limited resources for connectivity that is not only apparent in the mammalian neocortex but also in computer hardware or neuromorphic (brain-inspired) hardware by efficiently pruning synapses without losing performance.

Hippocampal Structure Predicts Statistical Learning and Associative Inference Abilities during Development

Despite the importance of learning and remembering across the lifespan, little is known about how the episodic memory system develops to support the extraction of associative structure from the environment. Here, we relate individual differences in volumes along the hippocampal long axis to performance on statistical learning and associative inference tasks-both of which require encoding associations that span multiple episodes-in a developmental sample ranging from ages 6 to 30 years. Relating age to volume, we found dissociable patterns across the hippocampal long axis, with opposite nonlinear volume changes in the head and body. These structural differences were paralleled by performance gains across the age range on both tasks, suggesting improvements in the cross-episode binding ability from childhood to adulthood. Controlling for age, we also found that smaller hippocampal heads were associated with superior behavioral performance on both tasks, consistent with this region's hypothesized role in forming generalized codes spanning events. Collectively, these results highlight the importance of examining hippocampal development as a function of position along the hippocampal axis and suggest that the hippocampal head is particularly important in encoding associative structure across development.