An Implantable Device to predict and Prevent Seizures

全世界有5000万人患有癫痫，仅在美国就有250万人。25%的复发性癫痫患者无法通过目前的药物或手术治疗来控制，他们必须求助于大剂量的镇静药物或实验性治疗。即使癫痫发作得到控制，患者也承受着神经系统和药物副作用的沉重负担。我们建议通过佐治亚理工学院和宾夕法尼亚大学，联合宾夕法尼亚大学、佐治亚理工学院、费城儿童医院和小型初创公司IntelliMedix的研究人员，在5到10年的时间里，集中精力开发一种治疗难治性癫痫的新疗法。一种植入式闭环系统，能够在电和行为发作之前预测癫痫发作，并在临床表现之前触发干预以中止癫痫发作。这个多样化的研究小组代表了多个学科和专业领域，包括生物工程、计算机科学、神经网络的计算建模、图像处理、临床成人和儿童癫痫、细胞和分子神经科学、神经生理学和神经药理学。这项工作将有三个主要重点：(1)癫痫发作预测：开发和完善癫痫发作预测算法，该算法来自于从成人、儿童和人类癫痫动物模型中植入的生物传感器获得的数据，能够在电和临床发作前几小时到几分钟预测癫痫发作；(2)癫痫发生机制；通过动物体外和体内实验、癫痫手术患者的记录，揭示这些算法所识别的观察到的信号变化背后的细胞、分子、神经生理和神经网络机制，并通过计算机模拟对这些发现进行建模，以完善预测和干预策略。如上所述，针对“致痫”过程中的特定点制定策略，包括电生理和药物干预，以不干扰正常脑功能的方式破坏导致癫痫发作的级联事件。这项工作将直接产生商业上可行的知识产权，包括：植入式生物传感器，微型生物相容电刺激和药物输注硬件，刺激范例，定制药物，用于信号采集的定制软件/硬件接口，处理和同步与驱动治疗干预的算法。希望在5-10年内实现闭环癫痫发作预测和预防装置，并显著改善癫痫患者的生活质量。

Epilepsy affects 50 million people worldwide, and 2.5 million in the United States alone. Fully 25 percent of those with recurrent seizures cannot be controlled by current medical or surgical treatment, and must resort to high doses of sedating medications or experimental therapy. Even when seizures are controlled, patients bear a significant burden of neurological and medication side effects. We propose to assemble an ensemble of accomplished investigators from the University of Pennsylvania, Georgia Institute of Technology, Children's Hospital of Philadelphia and IntelliMedix, a small start-up company through the GIT and Penn, in an intensive five to ten year effort to create a novel therapy for refractory epilepsy: an implantable closed loop system capable of predicting epileptic seizures prior to electrical and behavioral onset and triggering intervention to abort them before clinical expression. This diverse group of investigators represents multiple disciplines and areas of expertise including bioengineering, computer science, computational modeling of neuronal networks, image processing, clinical adult and pediatric epilepsy, cellular and molecular neuroscience, neurophysiology and neuropharmacology. The work will have three major thrusts: (1) Seizure Prediction: Developing and refining seizure prediction algorithms derived from data obtained from implanted biosensors in adults, children and in animal models of human epilepsy, capable of predicting seizures hours to minutes prior to electrical and clinical onset, (2) Mechanisms of ictogenesis: Unraveling the cellular, molecular, neurophysiologic and neuronal network mechanisms underlying the observed signal changes identified by these algorithms through in-vitro and in-vivo experiments in animals, recordings in human candidates for epilepsy surgery, and modeling these findings via computer simulations in order to refine predictive and intervention strategies, (3) Therapeutics: Developing strategies aimed at specific points in the "ictogenic" process, as discovered above, consisting of electrophysiological and pharmacological interventions to disrupt the cascade of events which lead to seizures, in ways which do not interfere with normal brain function. This work will directly give rise to commercially viable intellectual property including: implantable biosensors, miniaturized biocompatible electrical stimulation and drug infusion hardware, stimulation paradigms, customized pharmacologic agents, customized software/hardware interfaces for signal acquisition, processing and synchronization with algorithms for driving therapeutic interventions. It is hoped that a closed loop seizure prediction and prevention device will be implementable in a 5-10 year period and will significantly improve the quality of life of individuals with epilepsy.