Theoretical Models of Neural Development

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.

A Stay in Friedrich Bonhoeffer's Lab in Tubingen in the Mid-eighties

The main focus of research for which Friedrich Bonhoeffer's work is known in the Neuroscience community was pioneer experiments on how axonal projections could organize into "maps", what mechanisms are involved in axon guidance and involve gradients of guiding molecules, and isolation of the first such molecules, e.g. RAGS (ephrin A5) and RGM (repulsive guidance molecule). Other papers have described in detail these contributions as well as Friedrich Bonhoeffer's personality. In the mid-eighties, I made a 2-year stay in his lab and initiated a line of research on development of binocular connections in Mammals, particularly the guidance of retinal fibers to one or the other side of the brain. In this paper I recall these circumstances as they pertain to Neuroscience as it stood at the time, and explain as best as I can how his lab was a conducive setting for the discoveries made there and how Friedrich Bonhoeffer acted for me as a scientist and a tutor.

Mathematical models of neuronal growth

The establishment of a functioning neuronal network is a crucial step in neural development. During this process, neurons extend neurites—axons and dendrites—to meet other neurons and interconnect. Therefore, these neurites need to migrate, grow, branch and find the correct path to their target by processing sensory cues from their environment. These processes rely on many coupled biophysical effects including elasticity, viscosity, growth, active forces, chemical signaling, adhesion and cellular transport. Mathematical models offer a direct way to test hypotheses and understand the underlying mechanisms responsible for neuron development. Here, we critically review the main models of neurite growth and morphogenesis from a mathematical viewpoint. We present different models for growth, guidance and morphogenesis, with a particular emphasis on mechanics and mechanisms, and on simple mathematical models that can be partially treated analytically.

On the Role of LGN/V1 Spontaneous Activity as an Innate Learning Pattern for Visual Development

Correlated, spontaneous neural activity is known to play a necessary role in visual development, but the higher-order statistical structure of these coherent, amorphous patterns has only begun to emerge in the past decade. Several computational studies have demonstrated how this endogenous activity can be used to train a developing visual system. Models that generate spontaneous activity analogous to retinal waves have shown that these waves can serve as stimuli for efficient coding models of V1. This general strategy in development has one clear advantage: The same learning algorithm can be used both before and after eye-opening. This same insight can be applied to understanding LGN/V1 spontaneous activity. Although lateral geniculate nucleus (LGN) activity has been less discussed in the literature than retinal waves, here we argue that the waves found in the LGN have a number of properties that fill the role of a training pattern. We make the case that the role of “innate learning” with spontaneous activity is not only possible, but likely in later stages of visual development, and worth pursuing further using an efficient coding paradigm.

Technique Advances in Enteroatmospheric Fistula Isolation After Open Abdomen: A Review and Outlook

Enteroatmospheric fistula (EAF) after open abdomen adds difficulties to the management and increases the morbidity and mortality of patients. As an effective measurement, reconstructing gastrointestinal tract integrity not only reduces digestive juice wasting and wound contamination, but also allows expedient restoration of enteral nutrition and intestinal homeostasis. In this review, we introduce several technologies for the temporary isolation of EAF, including negative pressure wound therapy, fistuloclysis, fistula patch, surgical covered stent, three-dimensional (3D) printing stent, and injection molding stent. The manufacture and implantation procedures of each technique with their pros and cons are described in detail. Moreover, the approach in combination with finger measurement, x-ray imaging, and computerized tomography is used to measure anatomic parameters of fistula and design appropriate 3D printer-recognizable stereolithography files for production of isolation devices. Given the active roles that engineers playing in the technology development, we call on the cooperation between clinicians and engineers and the organization of clinical trials on these techniques.

BCNNM: A Framework for in silico Neural Tissue Development Modeling

Cerebral (“brain”) organoids are high-fidelity in vitro cellular models of the developing brain, which makes them one of the go-to methods to study isolated processes of tissue organization and its electrophysiological properties, allowing to collect invaluable data for in silico modeling neurodevelopmental processes. Complex computer models of biological systems supplement in vivo and in vitro experimentation and allow researchers to look at things that no laboratory study has access to, due to either technological or ethical limitations. In this paper, we present the Biological Cellular Neural Network Modeling (BCNNM) framework designed for building dynamic spatial models of neural tissue organization and basic stimulus dynamics. The BCNNM uses a convenient predicate description of sequences of biochemical reactions and can be used to run complex models of multi-layer neural network formation from a single initial stem cell. It involves processes such as proliferation of precursor cells and their differentiation into mature cell types, cell migration, axon and dendritic tree formation, axon pathfinding and synaptogenesis. The experiment described in this article demonstrates a creation of an in silico cerebral organoid-like structure, constituted of up to 1 million cells, which differentiate and self-organize into an interconnected system with four layers, where the spatial arrangement of layers and cells are consistent with the values of analogous parameters obtained from research on living tissues. Our in silico organoid contains axons and millions of synapses within and between the layers, and it comprises neurons with high density of connections (more than 10). In sum, the BCNNM is an easy-to-use and powerful framework for simulations of neural tissue development that provides a convenient way to design a variety of tractable in silico experiments.

Achieving functional neuronal dendrite structure through sequential stochastic growth and retraction

Class I ventral posterior dendritic arborisation (c1vpda) proprioceptive sensory neurons respond to contractions in the Drosophila larval body wall during crawling. Their dendritic branches run along the direction of contraction, possibly a functional requirement to maximise membrane curvature during crawling contractions. Although the molecular machinery of dendritic patterning in c1vpda has been extensively studied, the process leading to the precise elaboration of their comb-like shapes remains elusive. Here, to link dendrite shape with its proprioceptive role, we performed long-term, non-invasive, in vivo time-lapse imaging of c1vpda embryonic and larval morphogenesis to reveal a sequence of differentiation stages. We combined computer models and dendritic branch dynamics tracking to propose that distinct sequential phases of stochastic growth and retraction achieve efficient dendritic trees both in terms of wire and function. Our study shows how dendrite growth balances structure-function requirements, shedding new light on general principles of self-organisation in functionally specialised dendrites.

Distinct Microtubule Organizing Center Mechanisms Combine to Generate Neuron Polarity and Arbor Complexity

Dendrite and axon arbor wiring patterns determine the connectivity and computational characteristics of a neuron. The identities of these dendrite and axon arbors are created by differential polarization of their microtubule arrays, and their complexity and pattern are generated by the extension and organization of these arrays. We describe how several molecularly distinct microtubule organizing center (MTOC) mechanisms function during neuron differentiation to generate and arrange dendrite and axon microtubules. The temporal and spatial organization of these MTOCs generates, patterns, and diversifies arbor wiring.

Collective almost synchronization-based model to extract and predict features of EEG signals

Understanding the brain is important in the fields of science, medicine, and engineering. A promising approach to better understand the brain is through computing models. These models were adjusted to reproduce data collected from the brain. One of the most commonly used types of data in neuroscience comes from electroencephalography (EEG), which records the tiny voltages generated when neurons in the brain are activated. In this study, we propose a model based on complex networks of weakly connected dynamical systems (Hindmarsh–Rose neurons or Kuramoto oscillators), set to operate in a dynamic regime recognized as Collective Almost Synchronization (CAS). Our model not only successfully reproduces EEG data from both healthy and epileptic EEG signals, but it also predicts EEG features, the Hurst exponent, and the power spectrum. The proposed model is able to forecast EEG signals 5.76 s in the future. The average forecasting error was 9.22%. The random Kuramoto model produced the outstanding result for forecasting seizure EEG with an error of 11.21%.

Adaptive Landscape Shaped by Core Endogenous Network Coordinates Complex Early Progenitor Fate Commitments in Embryonic Pancreas

The classical development hierarchy of pancreatic cell fate commitments describes that multipotent progenitors (MPs) first bifurcate into tip cells and trunk cells, and then these cells give rise to acinar cells and endocrine/ductal cells separately. However, lineage tracings reveal that pancreatic progenitors are highly heterogeneous in tip and trunk domains in embryonic pancreas. The progenitor fate commitments from multipotency to unipotency during early pancreas development is insufficiently characterized. In pursuing a mechanistic understanding of the complexity in progenitor fate commitments, we construct a core endogenous network for pancreatic lineage decisions based on genetic regulations and quantified its intrinsic dynamic properties using dynamic modeling. The dynamics reveal a developmental landscape with high complexity that has not been clarified. Not only well-characterized pancreatic cells are reproduced, but also previously unrecognized progenitors-tip progenitor (TiP), trunk progenitor (TrP), later endocrine progenitor (LEP), and acinar progenitors (AciP/AciP2) are predicted. Further analyses show that TrP and LEP mediate endocrine lineage maturation, while TiP, AciP, AciP2 and TrP mediate acinar and ductal lineage maturation. The predicted cell fate commitments are validated by analyzing single-cell RNA sequencing (scRNA-seq) data. Significantly, this is the first time that a redefined hierarchy with detailed early pancreatic progenitor fate commitment is obtained.

Multi-context blind source separation by error-gated Hebbian rule

Animals need to adjust their inferences according to the context they are in. This is required for the multi-context blind source separation (BSS) task, where an agent needs to infer hidden sources from their context-dependent mixtures. The agent is expected to invert this mixing process for all contexts. Here, we show that a neural network that implements the error-gated Hebbian rule (EGHR) with sufficiently redundant sensory inputs can successfully learn this task. After training, the network can perform the multi-context BSS without further updating synapses, by retaining memories of all experienced contexts. This demonstrates an attractive use of the EGHR for dimensionality reduction by extracting low-dimensional sources across contexts. Finally, if there is a common feature shared across contexts, the EGHR can extract it and generalize the task to even inexperienced contexts. The results highlight the utility of the EGHR as a model for perceptual adaptation in animals.