Bioengineering Research Partnership in Brain Dynamics

脑动力学生物工程研究合作伙伴关系

• 批准号: 6752089
• 负责人: James Chris Sackellares
• 金额: $ 90.4万
• 依托单位: UNIVERSITY OF FLORIDA
• 依托单位国家: 美国
• 财政年份: 2001
• 资助国家: 美国
• 起止时间: 2001-07-19 至 2006-04-30
• 项目状态: 已结题
• 来源: https://reporter.nih.gov/project-details/6752089
• 关键词: benzodiazepines; bioengineering /biomedical engineering; brain disorder chemotherapy; brain disorder diagnosis; brain electrical activity; diagnosis design /evaluation; disease /therapy duration; electroencephalography; epilepsy; human subject; human therapy evaluation; laboratory mouse; mathematical model; model design /development; neuroimaging; nonhuman therapy evaluation; patient oriented research; rapid diagnosis

Epilepsy is a common neurological disorder that causes spontaneous recurrent seizures. In spite of major advances in pharmacology, neuroimaging, clinical neurophysiology, and neurosurgery, many patients remain disabled due to uncontrolled seizures. We propose to develop novel diagnostic and therapeutic tools, based on recent discoveries regarding dynamical mechanisms initiating epileptic seizures. We have found characteristic preictal dynamical changes, detectable in the electroencephalogram (EEG), preceding seizures by over 30 minutes (preictal transition, PT). More recently, other investigators have confirmed the presence of PDT. Our research indicates that the PT is demonstrable in the EEG in approximately 90 percent of seizures and that automated paradigms can be used to predict seizures. The potential to predict seizures in advance provides an opportunity to develop innovative diagnostic and therapeutic approaches. Our specific aims are: (1) Specific Aim 1. To continue the development of dynamic measures for the quantification of the spatiotemporal properties of the epileptic transition (years 1-3); (2) To develop specific pattern recognition algorithms for a seizure warning system (SWS) based upon the on-line features of the dynamical properties of brain electrical activity (years 1-4); To implement the dynamic features and pattern recognition algorithms in a SWS for on-line, real-time detection of the preictal dynamical transition (years 2-4); and (4) To evaluate the effects of therapeutic interventions during the preictal transition (years 1-5). The specific spatiotemporal patterns of the PT vary from seizure to seizure and patient to patient. Thus, a sensitive and reliable SWS will require sophisticated signal processing techniques. Dynamical measures will be augmented by other powerful analytic approaches, including multivariate time-series analysis, pattern recognition algorithms, and optimization techniques. To this end, we have gathered experts in signal processing, optimization, V.L.S.I., neurophysiology, neuroanatomy, epilepsy, and neurosurgery. The work will involve the coordination of several research sites throughout the University of Florida Campus including the Brain Dynamics Laboratory (Malcolm Randall V.A. Medical Center), Computer NeuroEngineering Laboratory College of Engineering), Center for Applied Optimization (College of Engineering), an In vitro Neurophysiology Research Laboratory (University of Florida Brain Institute), an In Vivo Neurophysiology Laboratory (Department of Pediatrics) and the Epilepsy Monitoring Laboratory (Shands Hospital). We anticipate that the proposed efforts will result in prototype diagnostic software and devices by the end of year 5. We also will obtain preliminary data that will be used for the design and testing of implantable devices that will activate pulsed therapeutic interventions during the preictal transition.

癫痫是一种常见的神经系统疾病，会引起自发性癫痫发作。 尽管药理学，神经影像学，临床神经生理学和神经外科方面取得了重大进展，但由于癫痫发作不受控制，许多患者仍然残疾。 我们建议根据最新的有关发起癫痫发作的动力学机制的发现，开发新的诊断和治疗工具。 我们发现在脑电图（EEG）中可检测到的特征性前动力学变化，在癫痫发作之前30分钟以上（前过渡，PT）。 最近，其他研究人员证实了PDT的存在。 我们的研究表明，在大约90％的癫痫发作中，在脑电图中可以证明PT，并且可以使用自动范式来预测癫痫发作。 预测癫痫发作的潜力为开发创新的诊断和治疗方法提供了机会。 我们的具体目的是：（1）特定目的1。继续开发动态度量，以定量癫痫转变的时空特性（1-3年）； （2）基于大脑电活动的动力学特性的在线特征（1-4年），开发了癫痫发警告系统（SWS）的特定模式识别算法；在SWS中实现动态特征和模式识别算法，以实现在线，实时检测前动态过渡的实时检测（2-4年）； （4）评估在过渡前过渡期间治疗干预措施的影响（1-5年）。 PT的特定时空模式从癫痫发作到癫痫发作和患者到患者不等。 因此，敏感且可靠的SWS将需要复杂的信号处理技术。 动态措施将通过其他强大的分析方法来增强，包括多元时间序列分析，模式识别算法和优化技术。 为此，我们收集了信号处理，优化，V.L.S.I.，神经生理学，神经解剖学，癫痫和神经外科手术方面的专家。 这项工作将涉及整个佛罗里达大学校园的几个研究站点的协调癫痫监测实验室（Shands医院）。 我们预计，提出的努力将在第5年底之前导致原型诊断软件和设备。我们还将获得初步数据，这些数据将用于植入设备的设计和测试，这些数据将在过渡期间激活脉冲治疗干预措施。