Informing deep neural networks by multiscale principles of neuromodulatory systems

Histamine in the neocortex: Towards integrating multiscale effectors

Histamine is a modulatory neurotransmitter, which has received relatively less attention in the central nervous system than other neurotransmitters. The functional role of histamine in the neocortex, the brain region that controls higher-order cognitive functions such as attention, learning and memory, remains largely unknown. This article focuses on the emerging roles and mechanisms of histamine release in the neocortex. We describe gaps in current knowledge and propose the application of interdisciplinary tools to dissect the detailed multiscale functional logic of histaminergic action in the neocortex ranging from sub-cellular, cellular, dendritic and synaptic levels to microcircuits and mesoscale effects.

Injectable Fluorescent Neural Interfaces for Cell-Specific Stimulating and Imaging

Building on current explorations in chronic optical neural interfaces, it is essential to address the risk of photothermal damage in traditional optogenetics. By focusing on calcium fluorescence for imaging rather than stimulation, injectable fluorescent neural interfaces significantly minimize photothermal damage and improve the accuracy of neuronal imaging. Key advancements including the use of injectable microelectronics for targeted electrical stimulation and their integration with cell-specific genetically encoded calcium indicators have been discussed. These injectable electronics that allow for post-treatment retrieval offer a minimally invasive solution, enhancing both usability and reliability. Furthermore, the integration of genetically encoded fluorescent calcium indicators with injectable bioelectronics enables precise neuronal recording and imaging of individual neurons. This shift not only minimizes risks such as photothermal conversion but also boosts safety, specificity, and effectiveness of neural imaging. Embracing these advancements represents a significant leap forward in biomedical engineering and neuroscience, paving the way for advanced brain-machine interfaces.

Task-Adaptive Neuromorphic Computing Using Reconfigurable Organic Neuristors with Tunable Plasticity and Logic-in-Memory Operations

The brain's function can be dynamically reconfigured through a unified neuron-synapse architecture, enabling task-adaptive network-level topology for energy-efficient learning and inferencing. Here, we demonstrate an organic neuristor utilizing a ferroelectric-electrolyte dielectric interface. This neuristor enables tunable short- to long-term plasticity and reconfigurable logic-in-memory functions by controlling the interfacial interaction between electrolyte ions and ferroelectric dipoles. Notably, the short-term plasticity of the organic neuristor allows for power-efficient reservoir computing in edge-computing scenarios, exhibiting impressive recognition accuracy, including images (90.6%) and acoustic signals (97.7%). For high-performance computing tasks, the neuristor based on long-term plasticity and logic-in-memory operations can construct all of the hardware circuits of a binarized neural network (BNN) within a unified framework. The BNN demonstrates excellent noise tolerance, achieving high recognition accuracies of 99.2% and 86.4% on the MNIST and CIFAR-10 data sets, respectively. Consequently, our research sheds light on the development of power-efficient artificial intelligence systems.

A brain-inspired algorithm that mitigates catastrophic forgetting of artificial and spiking neural networks with low computational cost

Neuromodulators in the brain act globally at many forms of synaptic plasticity, represented as metaplasticity, which is rarely considered by existing spiking (SNNs) and nonspiking artificial neural networks (ANNs). Here, we report an efficient brain-inspired computing algorithm for SNNs and ANNs, referred to here as neuromodulation-assisted credit assignment (NACA), which uses expectation signals to induce defined levels of neuromodulators to selective synapses, whereby the long-term synaptic potentiation and depression are modified in a nonlinear manner depending on the neuromodulator level. The NACA algorithm achieved high recognition accuracy with substantially reduced computational cost in learning spatial and temporal classification tasks. Notably, NACA was also verified as efficient for learning five different class continuous learning tasks with varying degrees of complexity, exhibiting a markedly mitigated catastrophic forgetting at low computational cost. Mapping synaptic weight changes showed that these benefits could be explained by the sparse and targeted synaptic modifications attributed to expectation-based global neuromodulation.

Intrinsic threshold plasticity: cholinergic activation and role in the neuronal recognition of incomplete input patterns

Activity-dependent changes in membrane excitability are observed in neurons across brain areas and represent a cell-autonomous form of plasticity (intrinsic plasticity; IP) that in itself does not involve alterations in synaptic strength (synaptic plasticity; SP). Non-homeostatic IP may play an essential role in learning, e.g. by changing the action potential threshold near the soma. A computational problem, however, arises from the implication that such amplification does not discriminate between synaptic inputs and therefore may reduce the resolution of input representation. Here, we investigate consequences of IP for the performance of an artificial neural network in (a) the discrimination of unknown input patterns and (b) the recognition of known/learned patterns. While negative changes in threshold potentials in the output layer indeed reduce its ability to discriminate patterns, they benefit the recognition of known but incompletely presented patterns. An analysis of thresholds and IP-induced threshold changes in published sets of physiological data obtained from whole-cell patch-clamp recordings from L2/3 pyramidal neurons in (a) the primary visual cortex (V1) of awake macaques and (b) the primary somatosensory cortex (S1) of mice in vitro, respectively, reveals a difference between resting and threshold potentials of ∼15 mV for V1 and ∼25 mV for S1, and a total plasticity range of ∼10 mV (S1). The most efficient activity pattern to lower threshold is paired cholinergic and electric activation. Our findings show that threshold reduction promotes a shift in neural coding strategies from accurate faithful representation to interpretative assignment of input patterns to learned object categories. KEY POINTS: Intrinsic plasticity may change the action potential threshold near the soma of neurons (threshold plasticity), thus altering the input-output function for all synaptic inputs 'upstream' of the plasticity location. A potential problem arising from this shared amplification is that it may reduce the ability to discriminate between different input patterns. Here, we assess the performance of an artificial neural network in the discrimination of unknown input patterns as well as the recognition of known patterns subsequent to changes in the spike threshold. We observe that negative changes in threshold potentials do reduce discrimination performance, but at the same time improve performance in an object recognition task, in particular when patterns are incompletely presented. Analysis of whole-cell patch-clamp recordings from pyramidal neurons in the primary somatosensory cortex (S1) of mice reveals that negative threshold changes preferentially result from electric stimulation of neurons paired with the activation of muscarinic acetylcholine receptors.

Distinctive properties of biological neural networks and recent advances in bottom-up approaches toward a better biologically plausible neural network

Although it may appear infeasible and impractical, building artificial intelligence (AI) using a bottom-up approach based on the understanding of neuroscience is straightforward. The lack of a generalized governing principle for biological neural networks (BNNs) forces us to address this problem by converting piecemeal information on the diverse features of neurons, synapses, and neural circuits into AI. In this review, we described recent attempts to build a biologically plausible neural network by following neuroscientifically similar strategies of neural network optimization or by implanting the outcome of the optimization, such as the properties of single computational units and the characteristics of the network architecture. In addition, we proposed a formalism of the relationship between the set of objectives that neural networks attempt to achieve, and neural network classes categorized by how closely their architectural features resemble those of BNN. This formalism is expected to define the potential roles of top-down and bottom-up approaches for building a biologically plausible neural network and offer a map helping the navigation of the gap between neuroscience and AI engineering.

Prefrontal Functional Connectivities in Autism Spectrum Disorders: A Connectopathic Disorder Affecting Movement, Interoception, and Cognition

The prefrontal cortex is included in a neuronal system that includes the basal ganglia, the thalamus, and the cerebellum. Most of the higher and more complex motor, cognitive, and emotional behavioral functions are thought to be found primarily in the frontal lobes. Insufficient connectivity between the medial prefrontal cortex (mPFC) and other regions of the brain that are distant from each other involved in top-down information processing rely on the global integration of data from multiple input sources and enhance low level perception processes (bottom-up information processing). The reduced deactivation in mPFC and in the rest of the Default Network during global task processing is consistent with the integrative modulatory role served by the mPFC. We stress the importance of understanding the degree to which sensory and movement anomalies in individuals with autism spectrum disorder (ASD) can contribute to social impairment. Further investigation on the neurobiological basis of sensory symptoms and its relationship to other clinical features found in ASD is required Treatment perhaps should not be first behaviorally based but rather based on facilitating sensory motor development.

Adolescent neurodevelopment and psychopathology: The interplay between adversity exposure and genetic risk for accelerated brain ageing

In adulthood, stress exposure and genetic risk heighten psychological vulnerability by accelerating neurobiological senescence. To investigate whether molecular and brain network maturation processes play a similar role in adolescence, we analysed genetic, as well as longitudinal task neuroimaging (inhibitory control, incentive processing) and early life adversity (i.e., material deprivation, violence) data from the Adolescent Brain and Cognitive Development study (N = 980, age range: 9-13 years). Genetic risk was estimated separately for Major Depressive Disorder (MDD) and Alzheimer's Disease (AD), two pathologies linked to stress exposure and allegedly sharing a causal connection (MDD-to-AD). Adversity and genetic risk for MDD/AD jointly predicted functional network segregation patterns suggestive of accelerated (GABA-linked) visual/attentional, but delayed (dopamine [D2]/glutamate [GLU5R]-linked) somatomotor/association system development. A positive relationship between brain maturation and psychopathology emerged only among the less vulnerable adolescents, thereby implying that normatively maladaptive neurodevelopmental alterations could foster adjustment among the more exposed and genetically more stress susceptible youths. Transcriptomic analyses suggested that sensitivity to stress may underpin the joint neurodevelopmental effect of adversity and genetic risk for MDD/AD, in line with the proposed role of negative emotionality as a precursor to AD, likely to account for the alleged causal impact of MDD on dementia onset.

Effects of neuromodulation-inspired mechanisms on the performance of deep neural networks in a spatial learning task

In recent years, the biological underpinnings of adaptive learning have been modeled, leading to faster model convergence and various behavioral benefits in tasks including spatial navigation and cue-reward association. Furthermore, studies have investigated how the neuromodulatory system, a major driver of synaptic plasticity and state-dependent changes in the brain neuronal activities, plays a role in training deep neural networks (DNNs). In this study, we extended previous studies on neuromodulation-inspired DNNs and explored the effects of neuromodulatory components on learning and single unit activities in a spatial learning task. Under the multiscale neuromodulatory framework, plastic components, dropout probability modulation, and learning rate decay were added to the single unit, layer, and whole network levels of DNN models, respectively. We observed behavioral benefits including faster learning and smaller error of ambulation. We then concluded that neuromodulatory components can affect learning trajectories, outcomes, and single unit activities, in a component- and hyperparameter-dependent manner.

Where is the error? Hierarchical predictive coding through dendritic error computation

Top-down feedback in cortex is critical for guiding sensory processing, which has prominently been formalized in the theory of hierarchical predictive coding (hPC). However, experimental evidence for error units, which are central to the theory, is inconclusive and it remains unclear how hPC can be implemented with spiking neurons. To address this, we connect hPC to existing work on efficient coding in balanced networks with lateral inhibition and predictive computation at apical dendrites. Together, this work points to an efficient implementation of hPC with spiking neurons, where prediction errors are computed not in separate units, but locally in dendritic compartments. We then discuss the correspondence of this model to experimentally observed connectivity patterns, plasticity, and dynamics in cortex.

Structural connections between the noradrenergic and cholinergic system shape the dynamics of functional brain networks

Complex cognitive abilities are thought to arise from the ability of the brain to adaptively reconfigure its internal network structure as a function of task demands. Recent work has suggested that this inherent flexibility may in part be conferred by the widespread projections of the ascending arousal systems. While the different components of the ascending arousal system are often studied in isolation, there are anatomical connections between neuromodulatory hubs that we hypothesise are crucial for mediating key features of adaptive network dynamics, such as the balance between integration and segregation. To test this hypothesis, we estimated the strength of structural connectivity between key hubs of the noradrenergic and cholinergic arousal systems (the locus coeruleus [LC] and nucleus basalis of Meynert [nbM], respectively). We then asked whether the strength of structural LC and nbM inter-connectivity was related to individual differences in the emergent, dynamical signatures of functional integration measured from resting state fMRI data, such as network and attractor topography. We observed a significant positive relationship between the strength of white-matter connections between the LC and nbM and the extent of network-level integration following BOLD signal peaks in LC relative to nbM activity. In addition, individuals with denser white-matter streamlines interconnecting neuromodulatory hubs also demonstrated a heightened ability to shift to novel brain states. These results suggest that individuals with stronger structural connectivity between the noradrenergic and cholinergic systems have a greater capacity to mediate the flexible network dynamics required to support complex, adaptive behaviour. Furthermore, our results highlight the underlying static features of the neuromodulatory hubs can impose some constraints on the dynamic features of the brain.