Cultured cortical networks described by conditional firing probabilities

Networks of cortical neurons were grown over multi electrode arrays to enable simultaneous measu-rement of action potentials from 60 electrodes. All possible pairs of electrodes (i,j) were tested for syn-chronized activity. We calculated conditional firing probability (CFPi,j[τ]) as the probability of an action potential at electrode j at t=τ, given that a spike was detected at i at t=0. If a CFPi,j[τ] distribution clearly deviated from flat, electrodes i and j were considered related. A function was fitted to each CFP-curve to obtain parameters for strength and delay. In young cultures the set of identified relationships changed rather quickly. At 16 days in vitro (DIV) 50% of the set changed within one day. Beyond 25 DIV this set stabilized: during a period of a week more than 50% of the set remained intact. Most individual relationships developed rather gradually. Moreover, beyond 25 DIV relational strength appeared quite stable during periods of ≈ 10 hours, with coefficients of variation (100×SD/mean) of ≈ 25% on average. CFP analysis provides a robust method to describe the stable underlying probabilistic structure of highly varying spontaneous activity in cultured cortical networks. It may offer a suitable basis for plasticity studies, in which induced changes should exceed spontaneous fluctuations. CFP analysis is likely to describe the network in sufficient detail to detect subtle changes in individual relationships. Analysis of data continuously recorded for ≈ 6 weeks, showed that highest stability is reached after ≈ 25 DIV, suggesting the 4th and 5th week as a suitable period for plasticity studies.