CAREER: Intelligent, Closed-Loop Neural Interfaces

神经接口将彻底改变神经系统疾病患者的疾病护理。如今，植入式神经刺激（NS）装置已被广泛应用于治疗运动障碍、疼痛和癫痫，并在治疗精神障碍、记忆丧失、抑郁症等方面显示出前景。临床神经刺激器有几个通道，提供简单的电脉冲，由医生编程，这个过程可能需要几个月甚至几年的时间。需要临床可行的、低功耗和小型化的系统，能够同时在多个通道上进行刺激和记录。提供同步传感和闭环控制将允许设备自动优化给定结果的参数，仅在出现症状时进行治疗，减少副作用并降低功率。闭合环路可以提供动态传递的疗法，根据患者的实时神经状态调整电刺激。此外，它们将首次使临床研究人员能够了解大脑对刺激的反应，并长期监测变化（可塑性）。未来的医疗保健以可穿戴和植入式设备为中心，这些设备将持续监测身体功能，并以智能、闭环和良好控制的方式自主治疗，而无需医生干预。这个未来需要广泛的、跨学科的工程，将电子学、人工智能和生物学结合起来。该项目的多学科性质延伸到促进下一代工程师动手生物电子科学和技术的长期教育目标。该项目开发了用于闭环神经调节的小型化和高度集成的设备，首次以真正的闭环方式将高通道计数神经记录与刺激结合起来。如果成功，拟议的工作将在两个领域取得重大进展。首先，将开发低功耗、低噪声集成电路设计技术，以结合同步刺激和记录。刺激会显著干扰神经记录，导致大量持续的伪影，掩盖或扭曲神经信号，模糊可靠的生物标志物检测。记录和刺激电路将共同设计，以消除这种相互作用。其次，在线机器学习算法将被开发并集成到硬件中，用于动态闭环控制。重点将是发展计算效率的技术，尽量减少电路的功耗。这两种技术将结合成一个单一的集成电路，用于多通道闭环神经调节，实现小占地和超低功耗，对人类长期使用是安全的。该装置将在动物疾病模型中进行测试。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

Neural interfaces will revolutionize disease care for patients of neurological conditions. Today, implantable neurostimulation (NS) devices have seen widespread adoption in the treatment of movement disorders, pain, and epilepsy, and have shown promise in treating psychiatric disorders, memory loss, depression, and more. Clinical neurostimulators have few channels and provide simple electrical pulses that are programmed by a doctor in a process that can take months or even years. There is a need for clinically viable, low-power and miniaturized systems that enable simultaneous stimulation and recording on many channels. Providing simultaneous sensing and closed-loop control will allow devices to automatically optimize parameters for a given outcome and only treat when symptoms are present, reducing side effects and reducing power. Closing the loop can provide dynamically delivered therapies that adjust electrical stimulation in response to a patient's real-time neural state. Furthermore, they will, for the first time, allow clinical researchers to understand the brain's response to stimulation and monitor changes (plasticity) over the long term. The future of medical care centers on wearable and implantable devices that will continuously monitor body functions and autonomously give treatment in an intelligent, closed-loop and well-controlled manner without the need for a doctor's intervention. This future requires broad, interdisciplinary engineering that combines electronics, artificial intelligence and biology. The multidisciplinary nature of this project extends to the long-term educational goals of promoting hands-on bioelectronic science and technology to the next generation of engineers. This project develops miniaturized and highly integrated devices for closed-loop neuromodulation that combine high channel count neural recording with stimulation in a truly closed-loop manner for the first time. If successful, the proposed work will make significant advancements in two areas. First, low-power, low-noise integrated circuit design techniques will be developed to combine simultaneous stimulation and recording. Stimulation can significantly interfere with neural recording, resulting in large, persistent artifacts that mask or distort the neural signal and obscure reliable biomarker detection. Recording and stimulation circuits will be co-designed to eliminate this interaction. Second, online machine learning algorithms will be developed and integrated in hardware for dynamic closed-loop control. The emphasis will be to develop computationally efficient techniques that minimize the power consumption of the circuit. The two techniques will be combined into a single integrated circuit for multi-channel closed-loop neuromodulation that achieves a small footprint and ultra-low power dissipation that is safe for chronic use in humans. The device will be tested in animal models of disease.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.