Detecting neuronal activity changes using an interspike interval algorithm compared with using visual inspection

A universally accepted method for efficiently detecting neuronal activity changes (NACs) in neurophysiological studies has not been established. Visual inspection is still considered to be one of the most reliable methods, although it is limited when it is used for analyzing large quantities of data. In this study, an algorithm that considers interspike intervals (ISIs) was developed to define the onset of NACs. Two criteria, involving the mean and the standard deviation (S.D.) of the ISIs during a control period, were used in the ISI algorithm to evaluate the NACs that occurred during a detection period. The first, an ISI decrease of more than 1 S.D. from the mean ISI of the control period, proved to be an effective criterion for qualifying the increased NACs (firing rate increases). The second, an ISI increase greater than 3 S.D.s, efficiently demarcated periods of decreased NACs (firing rate decreases). Statistically significant correlations between the detection of NAC onset times by the ISI algorithm and the detection of those times by visual inspections were observed after offline analyses of recorded neuronal activity. The present results suggest that this ISI algorithm is a reliable and efficient way of defining the onset of NACs.