A Distributed Neural Recorder/Stimulator for Epilepsy Monitoring

神经刺激改善生活质量的潜力现在已经得到了很好的证实，例如心脏起搏器（心律失常）、人工耳蜗（耳聋）、深部脑刺激器（帕金森病治疗）和慢性疼痛。到2024年，全球医疗神经调节市场将达到100亿美元。癫痫以异常和过度的大脑活动为特征，可导致剧烈癫痫发作，影响人口的0.5-1%（美国250万人，加拿大22.5万人）。多达30%的癫痫患者患有顽固性耐药癫痫，可从神经刺激中获益。最近的研究表明，癫痫发作预测在24小时内发生，为癫痫发作缓解措施和调整活动提供了机会。然而，长期的神经记录/刺激在多个皮层目标仍然具有挑战性。这种植入式装置最具挑战性的方面之一是在实用和临床安全的功率预算下，同时从分布在不同皮质部位的数千个电极获取和传输数据。该项目旨在开发一种针对癫痫的专用集成电路（ASIC），该集成电路将使神经记录/刺激设备具有单神经元分辨率、片上尖峰表征（振幅和形状）和刺激脉冲整形，同时使用代数编码方案无线传输压缩数据。该装置的目标植入物包括四个ASICS，植入大脑皮层适合癫痫发作预测的部位。这些专用集成电路连接到一个中央无线通信集线器，该集线器将压缩信息传输到患者佩戴的外部计算机。与其他癫痫记录器限制在几十个电极分布在几mm2的区域不同，每个集成电路小于1 mm2，并将同时记录来自1600个通道的20 μ m间距的峰值。与我们的商业合作伙伴合作，我们正在创造第一个同步和分布式6400电极神经记录/刺激系统，适合安全，终身植入。该项目是与Gezell（加拿大Gatineau），墨尔本大学（澳大利亚）及其工业合作伙伴Carbon Cybernetics Inc.（澳大利亚墨尔本）合作进行的，他们是非常高分辨率，100%生物相容性电极技术，超晶金刚石封装和癫痫预测算法的先驱。这一合作伙伴关系完美地补充了我们的团队组成的大学舍布鲁克（舍布鲁克，加拿大）和Gezell公司（加蒂诺，加拿大）在微电子和产品开发的专业知识。

The potential of neurostimulation to improve quality of life is now well established with examples such as Pacemakers (cardiac arrhythmia), cochlear implants (deafness), deep brain stimulators (Parkinson's treatment) and chronic pain. The global medical neuromodulation market is set to reach $10B by 2024.Epilepsy, characterized by abnormal and excessive brain activity which can lead to violent seizures, affects between 0.5-1% of the population (2.5 M people in United States, ~225k in Canada). As much as 30% of epilepsy patients suffer from intractable drug-resistant epilepsy and can benefit from neurostimulation. Recent research demonstrated seizure forecasting within 24 hours of occurrence, giving the opportunity for seizure mitigative measures and adjustment of activities. However, long-term neural recording/stimulation at multiple cortical targets is still challenging. Among the most challenging aspects of such an implantable device is the acquisition and transmission of data simultaneously from thousands of electrodes distributed at different cortical sites within a practical and clinically safe power budget. This project seeks to develop an application specific integrated circuit (ASIC) targeting epilepsy that will enable neurorecording/stimulation devices with single-neuron resolution, on-chip spike characterization (amplitude and shape) and stimulus pulse shaping, while wirelessly transmitting the compressed data using an algebraic coding scheme. The target implant for this device comprises four ASICS, implanted in the brain at cortical sites suitable for seizure prediction. These ASICs are wired to a central wireless communication hub that transmits the compressed information to an external computer worn by the patient. Unlike other epilepsy recorders limited to few tens of electrodes distributed on an area of few mm2, each integrated circuit is <1 mm2, and will simultaneously records spikes from 1600 channels at a 20 µm pitch. In collaboration with our commercial partners we are creating the first synchronized and distributed 6400-electrode neurorecorder/stimulator system suitable for safe, life-long implantation.This project is conducted in collaboration with Gezell (Gatineau, Canada), the University of Melbourne (Australia) and their industrial partner Carbon Cybernetics Inc. (Melbourne, Australia), who are pioneering very high resolution, 100% biocompatible electrode technology, ultrananocrystalline diamond enclosures and epilepsy prediction algorithms. This partnership perfectly complements our teams composed of the Université de Sherbrooke (Sherbrooke, Canada) and Gezell inc. (Gatineau, Canada) with expertise in microelectronics and product development.