Design Of A Neural Electrical Stimulator Based On Closed-Loop Control

Since the electrical stimulation has been widely used both in theoretic neurophysiologic research and therapeutic research of brain diseases such as epilepsy, the development of the electrical stimulators has made great progress. But the stimulators that able to generate stimulus according to the electrical activity recorded from the target area, i.e. brain, also called closed-loop stimulator, is very limited. Therefore, it is significant to design such a closed-loop control electrical stimulator either for physiologic research in laboratory or therapeutic treatments in clinical.In this thesis, we designed a LabVIEW-based closed-loop electrical stimulator, which mainly aimed at the rat's hippocampal field potential signals. The key hardware of this stimulator is composed from a data acquisition card (NI, USA) and an ordinary PC sound card. The main work and analyzing methods involved in the design of this stimulator were listed as followings:(1) Developed an algorithm to detect and predict the 9 rhythm among the field potentials of hippocampus. With this algorithm the stimulator can deliver stimulus right at the moment of the peak or trough point of 9 wave on the demand. (2) Designed a spike detection algorithm of epilepsy. This algorithm could help the stimulator generate stimulus after the emergence of epileptic spike with a pre-set delay. This function might be used in deep brain stimulation research. (3) Developed a simple epileptic EEG monitor with sound alarm. By taking advantage of a PC sound card Such vocal EEG monitor is cheap and easy to imitate.The main results of this thesis were summarized as followings:(1)The electrical stimulator can generate diverse stimulus waves selection, e.g., pulse, sine, triangle, sawtooth. It also supports alteration of stimulation settings, such as pulse width, stimulus interval, stimulus number, etc. The minimum pulse width is 50μs, meeting the need of most neurophysiological experiments.(2) The test results of the algorithms showed that average error between prediction ofθwave's peak and actual peak is 10% of period, while the average error of trough is 15%. The correct rates of peak and trough both are above 92%.(3) The results of epileptic spikes detection experiment showed that the epileptic detection algorithm could avoid the influence of normal EEG or noise signal and the accuracy of the detection of epileptic spikes could almost reach 100%.(4) The frequency response test of PC sound card showed that most of the sound card can meet the need of epilepsy EEG detection and the accuracy of the detection is around 97%.In a word, this paper designed a neural stimulator with closed-loop control, which can automatically predict the phase of 0 wave rhythm and detect epileptic spikes. It can also provide a variety of stimulus waves by users' set. Such stimulator can be used for the study of the mechanisms of nervous system or brain stimulation. At the same time, the simple epileptic EEG monitor by sound card has a high detection accuracy rate, which can be used in the field of telemedicine.