ASCENT: Ultra-high Throughput Neural Recording using Flexible, Polymer-based Shanks as Terahertz Dielectric Waveguides

脑机接口（BMI）用于记录大脑中神经元的电信号，并深入了解大脑和神经系统中发生的复杂过程。这种理解对于修复或增强认知和/或感觉和运动功能至关重要，这可能是必要的，例如，由于受伤或疾病对大脑造成的损害。传统的脑电图（EEG）或功能性磁共振成像（MRI）记录对于许多这些任务来说太粗糙，繁琐和缓慢;因此，需要使用植入微电极的BMI。然而，由刚性硅制成的最先进的可植入电极阵列（IEA）不仅寿命短，而且还可能损伤脑组织并导致疤痕形成。最近，IEA已经开发出使用柔性聚合物基柄，最大限度地减少植入期间的组织损伤，显着提高安全性，并为长期记录铺平道路。不幸的是，电极的数量以及因此可以记录的数据量非常有限。为了为神经科学的新基础科学发现和新的安全BMI的开发铺平道路，以治疗脑损伤或疾病，该项目引入了一种新的，完全无线的方法，具有几乎无限的数据带宽，用于在大脑外部传输数据，并通过生物相容性，柔性聚合物电极实现安全，长期的大脑记录。预计该系统将对推进IEA技术的最新发展产生巨大影响，因为它首次实现了长达数年的安全和高密度神经记录。硅-聚合物混合制造和通过聚合物波导进行芯片间通信的技术进步也将具有科学和实际应用价值。脑机接口面临的一个主要问题是在记录神经活动时实现高数据吞吐量和长寿命。为了克服目前对记录密度和寿命的限制，该项目将开发一种新的植入式电极阵列技术并制作原型，该技术将基于活性硅互补金属氧化物半导体（CMOS）的电极与生物相容性聚合物柄相结合。聚对二甲苯C聚合物是柔性的，并且可以被微制造，使得多个定制的、完全无线的CMOS神经记录小芯片可以沿着每个柄的长度沿着布置。聚合物柄作为电介质波导，将红光和太赫兹（THz）射频能量从大脑外部分别携带到小芯片，用于功率收集和反向散射数据通信，从而避免了对电线的需求。片上光电二极管将对入射光进行整流，为每个芯片供电，片上THz天线将用于通过反向散射通信将本地记录和放大的神经元数据调制到聚合物柄内的THz载波信号上。将在光学和太赫兹波长下对聚合物杆波导的电磁特性进行广泛建模，以优化杆横截面和太赫兹表面耦合天线的设计，从而最大限度地提高系统效率和通信带宽。每个小芯片将包含一个密集的神经记录电极阵列，该电极阵列具有有源放大、滤波和尖峰检测电路，用于在将神经元数据调制到THz载波信号上并通过聚合物柄波导将其发送到基座之前记录、数字化和压缩神经元数据。重要的是，这种范例实现了一个完全无线的系统，最大限度地增加了神经记录站点的数量，并贡献了一种新的混合硅聚合物架构，能够实现高效、高带宽的准光学芯片到芯片通信。该奖项反映了NSF的法定使命，并被认为值得通过使用基金会的智力价值和更广泛的影响审查标准进行评估来支持。

Brain Machine Interfaces (BMI) are used to record the electrical signals of the neurons in the brain and gain insights into the complex processes occurring in the brain and nervous system. This understanding is crucial to repair or augment cognitive and/or sensory and motor functions, which might be necessary, e.g., due to damage to the brain sustained by injuries or diseases. Traditional recording by electroencephalography (EEG) or functional magnetic resonance imaging (MRI) is too crude, cumbersome, and slow for many of these tasks; therefore, BMIs with implanted microelectrodes need to be used. However, state of the art implantable electrode arrays (IEAs) made from rigid silicon not only have short lifetimes but can also damage the brain tissue and cause scar formation. Recently, IEAs have been developed using flexible polymer-based shanks which minimize tissue damage during implantation, significantly increasing safety and paving the way towards long-term recording. Unfortunately, the number of electrodes, and thus the amount of data that can be recorded, is very limited. To pave the way for new basic science discoveries in neuroscience and the development of new, safe BMIs to treat individuals with brain injury or disease, this project introduces a new, completely wireless approach that has virtually unlimited data bandwidth for communicating data outside of the brain and enables safe, long term brain recording via biocompatible, flexible polymer electrodes. The system is expected to have a huge impact on advancing the state-of-the-art in IEA technology by enabling, for the first time ever, safe and high-density neural recording over multiple year-long durations. The technological advances in hybrid silicon-polymer fabrication and chip-to-chip communication via polymer waveguides will also hold scientific and practical application value in their own right.A major problem facing brain machine interfaces is achieving both high data throughput and long lifespans when recording neural activity. To overcome current limitations on recording density and lifetime, this project will develop and prototype a new implantable electrode array technology that combines active silicon complementary metal oxide semiconductor (CMOS)-based electrodes with a biocompatible polymer shank. The Parylene C polymer is flexible and can be microfabricated such that multiple custom, fully wireless CMOS neural recording chiplets can be arranged along the length of each shank. The polymer shank, acting as a dielectric waveguide, will carry both red light and Terahertz (THz) radio-frequency energy from outside the brain to the chiplets for power harvesting and backscatter data communication, respectively, obviating the need for wires. On-chip photodiodes will rectify the incident optical light for powering each chip, and on-chip THz antennas will be used to modulate the locally recorded and amplified neuronal data onto the THz carrier signal inside the polymer shank via backscatter communication. Extensive modeling will be done of the electromagnetic characteristics of the polymer shank waveguides at both optical and THz wavelengths in order to optimize the shank cross-section and design of the THz surface coupling antennas for maximizing system efficiency and communication bandwidth. Each chiplet will contain a dense neural recording electrode array with active amplification, filtering, and spike detection circuitry for recording, digitizing, and compressing neuronal data before modulating it on the THz carrier signal and sending it to the base via the polymer shank waveguide. Importantly, this paradigm achieves a completely wireless system that maximizes the number of neural recording sites and contributes a new hybrid silicon-polymer architecture capable of efficient, high-bandwidth quasi-optical chip-to-chip communication.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.