Characteristics Analysis Of Spontaneous Electrical Activity In Cultured Neuronal Network

Neuronal networks underlie memory storage and information processing in human brain. Investigating spontaneous activity of neuronal networks during individual development is very important in the understanding of the formations of neuronal circuits and their role in the plasticity and adaptability of neuronal network Multi-electrode arrays (MEA) is now a unique instrument in neuroscience research, which allows to observe the spatialtemporal propagating of neuronal activity networks.. Here, we used the MEA system to record the spontaneous networks activity in cultured hippocampal neuron, and analyzed these electrical signals investigated by MEA, Characteristics of spontaneous discharge of hippocampal neurons were analyzed, including the detection of burst, the characteristics of temporal coding, the complexity and the frequency, to reveal implications of these electric signals in the time and frequency domain.Bursts have been observed in cultured neuron networks in vitro, mammalian central nervous system and brain tissue sections. Bursts are regarded as important feature of spatiotemporal patterns in the neuronal firing. To improve the burst-detection self-adaptive algorithm, the parameter of the maximum inter-spike interval during the bursting is considered as a constraint condition to identify burst in this study. Experimental results indicated that the average validity for the burst detection was achieved about 93.8% by the improved method. Compared with the previous self-adaptive algorithm, the average validity was improved by 35.3%. This method was also used to investigate the spatiotemporal changes of spontaneous bursting activity in the development of cultured neuron networks. It was confirmed that bursts firing have specific characteristics in different developmental stages of neuronal networks. Furthermore, a mean inter-spike interval (MISI) method is described, which could be used to automatically detect bursts in spike trains. This method could auto-adaptively set a parameter according to the properties of the detected burst spike trains without artificially selecting or setting. Not only bursts can be automatically extracted from different bursting patterns, but also parameter changes of burst have been identified in neuron activity before and after electrical stimulus by this method.Regularities and characteristics of temporal coding in three typical patterns of the spontaneous neuron firing (bursting, continuous single spike firing, the single spike and complex spike bursts alternate firing pattern) were studied by adopting the interspike interval vs. time, interspike interval histograms (ISIH) and joint interspike interval (JISI). Moreover, the inter-spike interval vs. time was used to invesigate the temporal coding and information transmission in spike trains of spontaneous firing in cultured neuronal networks. These results implied that cultured neurons exhibit spontaneous and precisely repeating spike sequences.Approximate entropy (ApEn) was introduced to investigate the complexity and regularity in time-series data of different patterns of spontaneous neuron firing. The results suggested that the curves of dynamic changes in approximate entropy can predict and identify spikes and bursts of neuron firings, and can characterize changes of neuron activities in physiological state. This method was used to analyze the variety of information complexity in ISI time-series. The curves of dynamic changes in the complexity obviously present layered structures and small ApEn values for burst firing pattern. While the layered structures were not clear as complex spike bursts alternated with single spike in the firing patterns, and their ApEn values are larger than those of burst firing pattern. In addition, the complexity variety of spontaneous activity was studied in the network level by ApEn method. The results showed that the complexity of neuron activity rose like a wave with the development of neuronal network; the spatial structure of complexity of spontaneous neuron activity in the whole network was displayed by the ApEn gray-scale map.The time-frequency analysis methods, such as SPWVD, CWD and Morlet wavelet transform, were used to describe distribution of time-frequency of the spontaneous activity in the neuronal network. The results demonstrated that the information of the frequency changes with time for different patterns of spontaneous activities can be displayed by these time-frequency methods. The spontaneous activity is composed by multiple signals with different frequency, and the spikes firing are high-frequency signals. The frequency bands of high-energy for burst activities are lower than those of random single spike firing and continuous single spike firing.