本文關鍵詞：大鼠腦神經(jīng)元鋒電位誘發(fā)采集與分析軟件系統(tǒng)研究　出處：《新鄉(xiāng)醫(yī)學院》2015年碩士論文　論文類型：學位論文

  更多相關文章： 神經(jīng)接口 生物機器人 特征提取 spike分類

【摘要】：神經(jīng)接口(Neural Interface, NI)是連接外界環(huán)境與神經(jīng)系統(tǒng)的通道,外部世界的信息可以通過NI轉換成特定模式的電刺激輸入神經(jīng)系統(tǒng),用于改善、修復神經(jīng)系統(tǒng)的某些缺陷,也可以通過NI將神經(jīng)系統(tǒng)中的信息提取出來,經(jīng)過信號處理,翻譯成機器可以識別的命令來控制外部設備。生物機器人是體內植入了神經(jīng)接口的動物,通過神經(jīng)接口可以將外部控制命令轉化為電刺激輸入動物神經(jīng)系統(tǒng),控制動物的行為,同時也可以通過神經(jīng)接口將動物神經(jīng)系統(tǒng)中的神經(jīng)元電信號提取出來,經(jīng)過特征提取以及模式分類,可以觀察到動物的生理狀態(tài)。目前,應用于生物機器人的神經(jīng)接口大多數(shù)只是單向信息傳遞。單輸入型神經(jīng)接口無法定量評估刺激對動物的控制作用,單輸出型神經(jīng)接口由于缺乏刺激,無法控制動物的行為。因此,為了解決神經(jīng)激勵-響應模型、刺激參數(shù)的定量評價以及激勵參數(shù)的優(yōu)化等問題,實現(xiàn)對生物機器的有效精確控制,研究小型且具有雙向信息傳遞功能的神經(jīng)接口技術具有越來越大的研究價值。本文設計并搭建了基于神經(jīng)接口技術的生物機器人軟件系統(tǒng)平臺。本文首先對神經(jīng)接口技術和生物機器人進行了詳細的介紹,揭示了神經(jīng)接口技術研究的重要性和生物機器人研究的緊迫性,并提出了具體的研究內容。然后介紹了實現(xiàn)生物機器人的技術基礎,包括了電極制作、動物手術、Visual Studio 2010編程技術、Android智能手機開發(fā)平臺以及神經(jīng)元電信號處理技術。其次本文提出了基于雙向神經(jīng)接口的生物機器人系統(tǒng)框架,并詳細介紹和實現(xiàn)了系統(tǒng)中的刺激參數(shù)設置模塊、藍牙設備連接模塊、基于Android智能手機的刺激軟件、神經(jīng)元電信號解析模塊、實時信號采集模塊和神經(jīng)元信號處理模塊。刺激參數(shù)設置模塊和藍牙設備連接模塊實現(xiàn)了輸入型神經(jīng)接口,可以將實驗人員設置的命令翻譯成對應模式的電刺激輸入大鼠腦神經(jīng),控制大鼠行為：神經(jīng)元電信號解析模塊和神經(jīng)元信號處理實現(xiàn)了輸出型神經(jīng)接口,可以對大鼠神經(jīng)元發(fā)放的電信號進行采集以及分析處理。此外,本文設計實驗測試了生物機器人系統(tǒng)的性能。首先通過設計一個測試軟件驗證了輸入系統(tǒng)的連通性；其次,本文利用加噪后的模擬信號對算法性能進行測試；然后,本文利用改進后的算法對大鼠伸縮實驗范式下的腦神經(jīng)信號進行分類,得到了滿意的結果。最后,對全文進行總結,并指出在今后的科研工作中需要繼續(xù)努力研究的問題。
[Abstract]:Neural interface (NI) is a channel that connects the external environment with the nervous system. The information of the outside world can be transformed into a specific mode of electrical stimulation into the nervous system through NI, which can be used to improve and repair some defects of the nervous system, but also through NI to extract the information from the nervous system. After signal processing, it is translated into a machine-recognizable command to control an external device. A biological robot is an animal with a neural interface implanted in the body. Through the neural interface, the external control command can be transformed into electrical stimulation into the animal nervous system, and the behavior of the animal can be controlled. At the same time, the neural interface can also be used to extract the neuron electrical signals from the animal nervous system. After feature extraction and pattern classification, the physiological state of animals can be observed. Most of the neural interfaces used in biological robots are only one-way information transmission. Single input neural interface can not quantitatively evaluate the control effect of stimulation on animals, and single output neural interface is lack of stimulation. Therefore, in order to solve the problems of neural excitation-response model, quantitative evaluation of stimulation parameters and optimization of stimulation parameters, the effective and accurate control of biological machinery can be realized. It is more and more valuable to study the neural interface technology which has the function of bidirectional information transmission. This paper designs and builds the software system platform of the biological robot based on the neural interface technology. Interface technology and biological robot are introduced in detail. This paper reveals the importance of neural interface technology and the urgency of biological robot research, and puts forward the specific research contents. Then, it introduces the technical basis of realizing biological robot, including electrode fabrication. Visual Studio 2010 programming technique for animal surgery. Android smart phone development platform and neuron signal processing technology. Secondly, this paper proposes a bidirectional neural interface based biological robot system framework. And detailed introduction and implementation of the system stimulation parameter setting module, Bluetooth device connection module, based on Android smart phone stimulation software, neuron signal analysis module. The real-time signal acquisition module and neuron signal processing module, the stimulation parameter setting module and the Bluetooth device connection module realize the input neural interface. The commands set by the experimenter can be translated into electrical stimulation of corresponding mode into the rat brain nerve to control the behavior of the rats: the neuronal electrical signal analysis module and the neuronal signal processing implement the output neural interface. The electrical signals issued by rat neurons can be collected and analyzed. In this paper, experiments are designed to test the performance of the biological robot system. Firstly, a test software is designed to verify the connectivity of the input system. Secondly, the performance of the algorithm is tested by using the noised analog signal. Then, the improved algorithm is used to classify the neural signals in the rat extensional experiment paradigm, and the results are satisfactory. Finally, the paper summarizes the full text. And pointed out that in the future scientific research work need to continue to study the problems.
【學位授予單位】：新鄉(xiāng)醫(yī)學院
【學位級別】：碩士
【學位授予年份】：2015
【分類號】：TP242;R49

【參考文獻】

相關期刊論文 前2條

1 王斐;楊廣達;張丹;趙姝穎;;腦機接口在機器人控制中的應用研究現(xiàn)狀[J];機器人技術與應用;2012年06期

2 王克剛;齊麗英;;基于模糊C均值聚類和減法聚類結合的圖像分割[J];陜西理工學院學報(自然科學版);2008年02期

，

本文編號：1427714