CAREER:Bio-artificial Neuromorphic System Based on Synaptic Devices

职业：基于突触装置的生物人工神经形态系统

• 批准号: 1752241
• 负责人: Duygu Kuzum
• 金额: $ 50万
• 依托单位: University of California-San Diego
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2018
• 资助国家: 美国
• 起止时间: 2018-02-15 至 2023-01-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1752241&HistoricalAwards=false
• 关键词: CAREER; Bio; artificial; Neuromorphic; System

Engineering seamless interfaces between the brain and electronics is becoming increasingly important for advancing practical applications of new neurotechnologies and medical devices. To-date most of the efforts for interfacing with the brain have relied on microelectrode arrays. However, the interface between conventional microelectrodes and biological tissue is highly inefficient and unstable due to (i) Poor signal transduction between the tissue and electronics, (ii) Limited information transfer capacity of microelectrode arrays, and (iii) Severe foreign body response induced by invasive implantable microelectrodes. To bridge the gap between the brain and electronics, an interface technology, which can physically and functionally integrates with the biological tissue is needed. The aim of the proposed research is to develop a neuromorphic interface made of synthetic synaptic devices to form a stable, long-term input/output interface to the brain. Such a technology can help development of targeted and selective neuromodulation therapies for various neurological disorders (epilepsy, depression, memory disorders, etc) affecting one billion people worldwide. Long-term chronic studies enabled by this technology can revolutionize the speed of progress in brain activity mapping. The proposed interdisciplinary project will be integrated with educational and outreach activities, including hands-on laboratory training for students, rigorous training on fundamentals of bioelectronic devices for engineering specialists and outreach activities towards underrepresented communities from K-12 to graduate levels focusing on women and local Hispanic communities. Over the last decade, motivated partially by the race for developing better performing computation systems, significant progress has been made in developing neuromorphic devices and circuits that mimic biological neurons and synapses. So far, research on synaptic devices has only focused on increasing the efficiency of existing computational systems. This on-going research on neuromorphic devices, when combined with emerging neurotechnologies, offers unprecedented potential to engineer bio-artificial neuromorphic circuits, which can address the shortcomings of the brain computer interfaces and bridge the gap between the brain and electronics. The primary goal of the proposed research is to develop to a neuromorphic interface, which will serve as a translator adapting time, amplitude and shape characteristics of the electrical stimuli transmitted to/from the brain. The challenges that will be addressed during the course of the project include: (1) Development of biocompatible plastic synaptic devices, (2) Enhancement of synaptic device-cell coupling and effective signal transduction, and (3) Development of a platform for construction of bio-artificial neuromorphic tissue. In vivo integration of the neuromorphic tissue with the brain will also be investigated to demonstrate a stable neural interface with natural signal processing capability. The proposed work will build upon the PI's interdisciplinary expertise and experience in neuromorphic devices and implantable neural interfaces. The bio-artificial neuromorphic interface will be a key enabler to functionally transmit information to/from the brain overcoming the biocompatibility and biostability limitations of implantable devices.

设计大脑和电子设备之间的无缝接口对于推进新的神经技术和医疗设备的实际应用正变得越来越重要。到目前为止，大多数与大脑接口的努力都依赖于微电极阵列。然而，传统的微电极与生物组织之间的界面是非常低效和不稳定的，这是由于(I)组织与电子之间的信号转导不佳，(Ii)微电极阵列的信息传输能力有限，以及(Iii)侵入性可植入微电极引起的严重的异物反应。为了弥合大脑和电子设备之间的鸿沟，需要一种能够在物理上和功能上与生物组织集成的接口技术。这项拟议研究的目的是开发一种由合成突触设备组成的神经形态接口，以形成稳定的、长期的大脑输入/输出接口。这种技术可以帮助开发针对各种神经疾病(癫痫、抑郁症、记忆障碍等)的有针对性和选择性的神经调节疗法，这些疾病影响着全球10亿人。这项技术带来的长期慢性研究可以彻底改变大脑活动图谱的进展速度。拟议的跨学科项目将与教育和外联活动相结合，包括为学生提供动手实验室培训，为工程专家提供关于生物电子设备基础知识的严格培训，以及面向从K-12到研究生级别的代表性不足社区的外联活动，重点是妇女和当地拉美裔社区。在过去的十年里，部分受到开发性能更好的计算系统的竞赛的推动，在开发模仿生物神经元和突触的神经形态设备和电路方面取得了重大进展。到目前为止，对突触设备的研究只集中在提高现有计算系统的效率上。这项正在进行的神经形态设备研究，与新兴的神经技术相结合，为设计生物人工神经形态电路提供了前所未有的潜力，可以解决脑计算机接口的缺点，并弥合大脑和电子设备之间的差距。这项拟议研究的主要目标是发展成一个神经形态界面，它将作为一个翻译器，适应传入/传出大脑的电刺激的时间、幅度和形状特征。该项目将解决的挑战包括：(1)开发生物兼容的塑料突触装置，(2)加强突触装置-细胞耦合和有效的信号转导，以及(3)开发构建生物人工神经形态组织的平台。神经形态组织与大脑的体内整合也将被研究，以展示具有自然信号处理能力的稳定的神经接口。拟议的工作将建立在PI在神经形态设备和可植入神经接口方面的跨学科专业知识和经验的基础上。生物-人工神经形态接口将是从功能上向/从大脑传输信息的关键使能器，它克服了可植入设备的生物兼容性和生物稳定性限制。

项目成果

Adaptive Quantization as a Device-Algorithm Co-Design Approach to Improve the Performance of In-Memory Unsupervised Learning With SNNs

• DOI: 10.1109/ted.2019.2898402
• 发表时间: 2019-02
• 期刊: IEEE Transactions on Electron Devices
• 影响因子: 3.1
• 作者: Yuhan Shi;Zhisheng Huang;Sangheon Oh;Nathan Kaslan;Jungwoo Song;D. Kuzum

Multimodal neural recordings with Neuro-FITM uncover diverse patterns of cortical-hippocampal interactions.

• DOI: 10.1038/s41593-021-00841-5
• 发表时间: 2021-06
• 期刊: Nature neuroscience
• 影响因子: 25
• 作者: Liu X;Ren C;Lu Y;Liu Y;Kim JH;Leutgeb S;Komiyama T;Kuzum D
• 通讯作者: Kuzum D

Neuroinspired unsupervised learning and pruning with subquantum CBRAM arrays

• DOI: 10.1038/s41467-018-07682-0
• 发表时间: 2018-12-14
• 期刊: NATURE COMMUNICATIONS
• 影响因子: 16.6
• 作者: Shi, Yuhan;Nguyen, Leon;Kuzum, Duygu
• 通讯作者: Kuzum, Duygu

A Soft-Pruning Method Applied During Training of Spiking Neural Networks for In-memory Computing Applications

• DOI: 10.3389/fnins.2019.00405
• 发表时间: 2019-04-26
• 期刊: FRONTIERS IN NEUROSCIENCE
• 影响因子: 4.3
• 作者: Shi, Yuhan;Nguyen, Leon;Kuzum, Duygu
• 通讯作者: Kuzum, Duygu

A Neuromorphic Brain Interface Based on RRAM Crossbar Arrays for High Throughput Real-Time Spike Sorting

基于 RRAM 交叉阵列的神经形态脑接口，用于高通量实时尖峰排序

• DOI: 10.1109/ted.2021.3131116
• 发表时间: 2022
• 期刊: IEEE Transactions on Electron Devices
• 影响因子: 3.1
• 作者: Shi, Yuhan;Ananthakrishnan, Akshay;Oh, Sangheon;Liu, Xin;Hota, Gopabandhu;Cauwenberghs, Gert;Kuzum, Duygu
• 通讯作者: Kuzum, Duygu