Measurements Of Deterministic And Complexity Of Interspike Interval From Firing Neurons

Action Potential is one of the most important media for neurons to transmit information. Neurons keep touch with each other by sending series of action potential. To study activity of neurons ,firing frequency has been observed for a long time. But we can not understand more knowledge about the communication of neurons because firing frequency contains little information. At the same time, ISI(interspike interval) reflects the arrangement of action potential in time space,and there may be more information in the ISI data than that in firing frequency. We found that lots of ISI series observed in our lab looked like noise without rule. Is the ISI series noise or is there any inner rules in it9 How to determine the degree of complexity? These two questions puzzled us. And a set of new system which can sample electric signals of animals including action potential, firing frequency and ISI in real time.With the development of nonlinear dynamic theory, more and more nonlinear method are used to study the activity of neural system. Time series analysis become one of the most important methods to study neuronal codes. In this paper, we studied the deterministic mechanism of ISI with the unstableperiod orbit(UPO); we tested many ISI data from different model with the method of approximate entropy(ApEn) to study the method which is used in testing the degree of complicacy of ISI and we built up a system including hardware and software for our lab to view and record action potential, firing frequency and ISI in real time.Results1. Spontaneous firing patterns were observed from neurons of supraoptic nucleus (SON) in hypothalamus with the method of patch clamp. And the data of ISI(interspike interval) were taken out.2. We studied the ISI by the method of return map and any obvious geometrical structure was not found.1. We extract period 1 UPO from ISI data of SON. Then we studied it's local dynamics and a saddle-nod was found.4. We further studied the data and found period 2,3 UPO.5. With the method of ApEn , we calculated the ApEn of bifurcate data got from model of R-H. We found the ApEn of chaos is obviously higher than that of period 2 and 3.6.The ISI data from Bennett and Xie's model of injured nerve were calculatedand the result of data including period 2, chaos and period 3 is similar to that ofR-H model. 7.The ISI data of injured nerve with period 1 were calculated , but the ApEn isnot very good. We solved the problem by replacing window-SD with full-data-SD. 8. ApEn changed between 0 and 1 in ISI data from the spontaneous firing in SONneurons recorded.9.The action potential , firing frequency and ISI data by sampling system designed reflected the fact of our experiment.Conclusion1. The method of UPO can test the deterministic mechanism of interspick interval of neuron.2. The irregular ISI data of spontaneous firing from SON is not noise , there is a deterministic mechanism in it.3. ApEn can test the degree of complexity degree of ISI . Some changes of the method can be done to make the result reasonable while there is special data.4. The sampling system designed can completely view and record action potential, firing frequency and ISI in real time .