LFPs coding working memory task via information entropy using plugin method / 国际生物医学工程杂志

ABSTRACT

Objective Toinvestigatetheentropyoflocalfieldpotentials(LFPs)recordedinratmedialprefrontal cortex during a Y-maze working memory (WM) task, to provide computing support for neural coding mechanism.Methods Sixteen-channel LFPs were recorded from SD rats while they performed a Y-maze WM task.The data came from 4 rats, 20 trials (10 correct trials and 10 incorrect trials) per rat provided by laboratory of neurobiology in medicine,Tianjin Medical University.Original LFPs were preprocessed to remove 50 Hz power line noise and baseline drift.Multi-taper Fourier transform was applied to calculate spatial distributions of LFPs and band pass filter were used to extract characteristic signal.The entroy coding of 16 channel LFPs was as follows the physiological window was set to be 500 ms, the step length of physiological window was set to be 125 ms, windows were added to LFPs data, and then LFPs entropy of each sliding window was computed and averaged to get the trend of multichannel entropy values duringthe WM task.Results The power of θ band (4-12 Hz) in LFPs increased.The averaged entropy value ofmultichannel θ band LFPs in correct trials was 0.939±-0.020, which were larger than those in the resting state, 0.795±0.031 (P＜0.05).Those during wrong WM task had no significant difference, which didn't encode the WM task.Conclusions The principal frequency band related to WM is the θ band and LFPs entropy encodes the WM effectively.