An Informational Approach to the Analysis of the Low-Frequency Neuronal Impulse Activity in the Rostral Hypothalamus

Statistical analysis of the low-frequency (1.0 sec^-1 and lower) neuronal impulse activity (IA) meets a few fundamental difficulties. Among them, the most significant is the small number of measurements (interspike intervals) recorded within an acceptable analysis epoch. In our study, we examined the possibility of using the normalized (by its maximum value) informational entropy (Hn) for estimation of the significance of changes in the IA generated by low-frequency neurons of the rostral hypothalamus after electrical stimulation of the prefrontal cortex. We compared the efficiencies of using the U-test (Kolmogorov–Mann–Whitney) and Hn estimate for the analysis of the same samples of neuronal responses. The results allow us to conclude that Hn is a significantly more acceptable estimate for detection of stimulation-induced modifications of the IA generated by low-frequency neurons, as compared with the U-test. The direction of shifts in the Hn value makes it possible to estimate the pattern of neuronal response. This value reflects the state of the neuron and correlates with the type of neuronal responses.