EN
Local fi eld potential (LFP) – the result of summed postsynaptic potentials from cell populations – reveal the most characteristic neural activity at the recording site in the brain. Thus LFPs are well suited for study of neural networks, specifi cally those involved in processing of sensory information. We propose a method of assessing functional brain connectivity through LFP analysis. The method is applied to multi-site signals representing potentials evoked by a repeated, stereotyped stimulus. In spite of a stereotyped stimulus, all responses are different due to, inter alia, ongoing background activity of the brain and this trial-to-trial variability is utilized in our analysis. The method is based on calculation of correlations between trial-to-trial LFP variations at every post-stimulus latency at every recoding site. The results show how neuronal activities at different sites and latencies correspond to activation at other sites with a given time delays. We used this method to analyze the functional connectivity in thalamocortical network involved in processing of somatosensory (vibrissal) information in non-anaesthetised rat. One result is that the cortical activation at 25–50 ms post-stimulus correlates with thalamic LFP measured at 50–150 ms post stimulus, thus implying this late latency thalamic activity depends on a corticothalamic feedback. Apart from raw LFP, the method is applicable to results of various decomposition methods of brain signals (PCA, ICA, etc.)