Developing novel neural network tools for accurate and interpretable dynamical modeling of neural circuits

开发新型神经网络工具，用于准确且可解释的神经回路动态建模

• 批准号: 10752956
• 负责人: Christopher Versteeg
• 金额: $ 7.66万
• 依托单位: EMORY UNIVERSITY
• 依托单位国家: 美国
• 财政年份: 2023
• 资助国家: 美国
• 起止时间: 2023-08-01 至 2026-07-31
• 项目状态: 未结题
• 来源: https://reporter.nih.gov/project-details/10752956
• 关键词: Address; Affect; Architecture; Area; Behavior; Biological; Brain; Communication; Complex; Computer Models; Coupled; Coupling; Data; Data Set; Decision Making; Differential Equation; Dimensions; Dorsal; Electrodes; Generations; Geometry; Goals; Human; Individual; Intuition; Investigation; Learning; Measurement; Methodology; Methods; Modeling; Monkeys; Motor; Motor Cortex; Movement; Neural Network Simulation; Neurons; Neurosciences; Output; Pattern; Performance; Population; Publishing; Rewards; Speed; System; Systems Theory; Testing; Time; Training; Variant; artificial neural network; autoencoder; biological systems; cognitive process; dynamic system; frontal lobe; in silico; innovation; insight; invention; multidimensional data; neural; neural circuit; neural model; neural network; neural network architecture; neuromechanism; neuronal patterning; novel; reconstruction; recurrent neural network; terabyte; theories; time interval; tool; Address; Affect; Architecture; Area; Behavior; Biological; Brain; Communication; Complex; Computer Models; Coupled; Coupling; Data; Data Set; Decision Making; Differential Equation; Dimensions; Dorsal; Electrodes; Generations; Geometry; Goals; Human; Individual; Intuition; Investigation; Learning; Measurement; Methodology; Methods; Modeling; Monkeys; Motor; Motor Cortex; Movement; Neural Network Simulation; Neurons; Neurosciences; Output; Pattern; Performance; Population; Publishing; Rewards; Speed; System; Systems Theory; Testing; Time; Training; Variant; artificial neural network; autoencoder; biological systems; cognitive process; dynamic system; frontal lobe; in silico; innovation; insight; invention; multidimensional data; neural; neural circuit; neural model; neural network; neural network architecture; neuromechanism; neuronal patterning; novel; reconstruction; recurrent neural network; terabyte; theories; time interval; tool

Abstract In recent years, the number of neurons that we can record simultaneously has seen an exponential increase, presenting a daunting challenge: how do we analyze these complex and high-dimensional datasets to gain insight into how neural circuits perform computation? Tools from dynamical systems theory have successfully unraveled the computational machinery of artificial recurrent neural networks (RNNs) trained to perform goal-directed tasks. If we could apply these tools to biological neural circuits, it would provide unparalleled access to the inner workings of the brain and potentially allow us to connect theories of neural computation to real biological data. However, for these tools to be useful, we need to create in silico replicas whose dynamics faithfully represent the dynamics of the underlying biological system. To date, the best in silico replicas of biological networks are RNNs trained to produce output that matches recorded patterns of neuronal firing. While this approach is rapidly growing in popularity, it has critical flaws. Current training methodologies are not constrained to produce accurate representations of the underlying dynamics; in fact, RNNs are actually rewarded for inventing superfluous dynamics, so long as those dynamics help to reproduce recorded neural data. Additionally, these models often assume that the relationship (“embedding”) between latent activity and neural firing rates is linear; when this assumption proves false, the dynamical accuracy suffers. The problems of superfluous dynamics and non-linear embedding are especially severe when attempting to model a system of interacting neural circuits. The objective of this proposal is to develop a novel artificial neural network architecture that addresses the above challenges and allows our in-silico models to capture accurate dynamics that are built both within and across-circuits. My approach combines two key components: 1) neural ordinary differential equations (NODEs), a computational architecture that we have demonstrated learns dynamics more accurately and compactly than RNNs and 2) invertible neural network (INN) readouts, which eliminate superfluous dynamics and allow the model to approximate nonlinear embeddings. I will validate the ability of this model, called an Ordinary Differential equation auto-encoder with Invertible readout (ODIN), to find accurate within- and across-circuit dynamics using synthetic neural data and previously-collected multi-electrode recordings from monkeys. This tool will help to build a bridge between neural data and both local and distributed neural computations.

抽象的 近年来，我们可以记录的神经元数量已经达到了指数 增加，提出艰巨的挑战：我们如何分析这些复杂且高维 数据集可深入了解神经回路如何执行计算？动态系统的工具 理论成功地揭示了人工复发神经网络的计算机制 （RNN）经过培训，可以执行目标指导的任务。如果我们可以将这些工具应用于生物神经回路， 它将提供无与伦比的大脑内部运作的访问，并有可能使我们能够 将神经功能的理论与真实的生物学数据联系起来。但是，对于这些工具有用， 我们需要在计算机复制品中创建，其动态忠实地代表了基础的动态 生物系统。 迄今 匹配记录了神经元射击的模式。虽然这种方法正在迅速增长，但它具有 关键缺陷。当前的培训方法不受限制地产生准确的表示 基础动力学；实际上，RNN实际上因发明多余的动态而获得了回报，因此 只要这些动力有助于再现记录的神经数据。另外，这些模型经常 假设潜在活动和神经发射率之间的关系（“嵌入”）是线性的。什么时候 该假设证明是错误的，动态精度会受到损失。多余的动态问题 在尝试建模相互作用的系统时，非线性嵌入特别严重 神经回路。 该建议的目的是开发一种新颖的人工神经网络架构 解决上述挑战，并允许我们的内部模型捕获准确的动态 在跨电路内部和跨电路中建造。我的方法结合了两个关键组成部分：1）神经普通 微分方程（节点），我们已经证明学习的计算体系结构 比RNN和2）可逆神经网络（INN）读数更准确，更紧凑， 消除了多余的动力学，并允许模型近似非线性嵌入。我会 验证该模型的能力，称为普通微分方程自动编码器，具有可逆性 读数（ODIN），使用合成神经数据和 先前收集的猴子的多电极记录。该工具将有助于建造桥梁 在神经元数据以及局部和分布式神经元计算之间。