• 主题
高级搜索  >
历史搜索 清空历史
首页> 外文会议>Design of Medical Devices Conference >NEUROMORPHIC REPRESENTATION OF CARDIAC DATA FROM THE AMERICAN BLACK BEAR DURING HIBERNATION
【24h】

NEUROMORPHIC REPRESENTATION OF CARDIAC DATA FROM THE AMERICAN BLACK BEAR DURING HIBERNATION

机译：在冬眠期间美国黑熊的心脏数据的神经形态表示

获取原文
页面导航
  • 摘要
  • 著录项
  • 相似文献
  • 相关主题

摘要

Brain-inspired (neuromorphic) systems realize biological neural principles with Spiking Neural Networks (SNN) to provide high-performing, energy-efficient frameworks for robotics, artificial intelligence, and adaptive control. The Neural Engineering Framework (NEF) brings forth a theoretical framework approach for the representation of high-dimensional mathematical constructs with spiking neurons for the implementation of functional large-scale neural networks. Here, we explore the utilization of neuromorphic adaptive control for circadian modulated cardiac pacing by examining the neuromorphic representation of high-dimensional cardiac data. For this study, we have utilized a model from a data set acquired from an American black bear during hibernation. Black bears in Minnesota will hibernate for 4-6 months without eating and drinking while losing little muscle mass and remain relatively normothermic throughout the winter [10]. In the current study, we obtained EEG and ECG data from one black bear throughout the winter months in Grand Rapids, MN;, represented with NEF. Our results demonstrated opposing requirements for neuromorphic representation. While using high synaptic time constants for obtained ECG data, provided desirable low pass filtering, representation of EEG data requires fast synapses and a high number of neurons. Although this is only an analysis of a small sample of the data available, these guidelines provided the robust pilot dataset to observe the SNN patterns during prolonged hibernation and pair this data with the cardiac responses and thus support research questions related to the autonomic tone during hibernation. This preliminary research will help further develop our neuromorphic adaptive controller to better adapt cardiac pacing to circadian rhythms. This unique dataset may pave the way toward deciphering the underlying neural mechanisms of hibernation, providing translational to humans.
机译：脑激发（神经形态）系统实现了尖刺神经网络（SNN）的生物神经原理，为机器人，人工智能和自适应控制提供高性能，节能框架。神经工程框架（NEF）提出了一种具有尖峰神经元的高维数学构造的理论框架方法，用于实现功能大规模神经网络。在此，我们通过检查高维心脏数据的神经形态表示来探讨昼夜调节心脏起搏的神经形态适应性控制。对于这项研究，我们已经利用了从美国黑熊期间从美国黑熊期间获取的数据集的模型。明尼苏达州的黑熊将冬眠4-6个月，而不会饮酒，同时在冬季失去较少的肌肉质量并保持相对较高的常温[10]。在目前的研究中，我们在整个冬季的冬季，在大急流，Mn，Mn;与Nef表示的冬季，获得EEG和ECG数据。我们的结果表明了神经形态表现的反对要求。在使用获得的ECG数据的高突触时间常数的同时，提供所需的低通滤波，EEG数据的表示需要快速突触和大量神经元。虽然这只是对可用数据的小样本的分析，但是这些指南提供了强大的导频数据集，以便在长期休眠期间观察SNN模式并将这种数据与心脏反应配对，从而支持与自主语调期间的冬眠期相关的研究问题。这项初步研究将有助于进一步开发我们的神经形态自适应控制器，以更好地适应心脏起搏到昼夜节律。这个独特的数据集可以铺设朝向休眠的基础神经机制来铺平道路，为人类提供平移。

著录项

相似文献

  • 外文文献
  • 中文文献
  • 专利
获取原文

客服邮箱：kefu@zhangqiaokeyan.com

京公网安备：11010802029741号 ICP备案号：京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有

  • 客服微信

  • 服务号