Multi-channel Local Field Potentials-Spikes Spectrum Coherence Coding In Rat Prefrontal Cortex During Working Memory

ObjectiveWorking memory in brain is coded by different modal neural signals collaboratively, which is crossover study of neuroscience, cognitive science, information science area. This thesis study on working memory task in rats brain (prefrontal cortex) in vivo multichannel neural signals of two different modes: continuous time series form of multi-channel local field potentials (LFPs), discrete points of multi channel Spike (Spikes), and its code on working memory events. Papers of LFPs-Spikes dynamic spectrum coherent (DSC) coding method for its collaborative coding mode of working memory events, supporting for the study of Animal experiments and neural coding mechanisms of working memory.Methods1. Animal experimentsExperimental animals from the Chinese Academy of Medical Sciences Institute of radiation medicine experimental animal Center are SD rats 20. Base on rats in vivo multi-channel micro-electrode array acquisition techniques, use Cerebus-128 neural signal process system, recorde 16-channel neural signals in rat prefrontal cortex in Y-maze during working memory tasks.1) Rats working memory task training in Y-mazeWorking memory training in accordance with the specification at home and abroad:train 2 times a day,1 times 10 task, one task includes one free choice and one invert selection on the direction of Y-maze, timing task completion times using the stopwatch, two choice interval 5-10s. Training continued until the rat consecutive 2 days the correct rate of working memory task is greater than 80%, which had already been formed working memory.2) Multi-channel micro-electrode implantation in vivoFixed in rats after anesthesia on stereotaxic instrument, select the prefrontal cortex (2.5-4.5mm anterior to bragma and 0.2-1.0mm lateral to midline) opened a rectangular window, implanted 16 micro-electrode array, when it reaches the target depth 2.5-3mm deep from cortical surface, with dental cement. Postoperative recovery for 5-7 days.3) Collect multi-channel neural signals in working memory task of ratsApplication Cerebus-128 multi-channel neural signal process system records 16 channel neural signals in rats of prefrontal cortex in vivo during working memory tasks in Y-maze. Each test records several complete working memory tasks.2. Multi-channel raw data pre-processingMulti-channel LFPs were acquired through lowpass filtering the raw data (0.3-500Hz). Local linear regression based on the weighted least square method was used to remove the baseline drift and power-line interference on LFPs. Get zero mean multi-channel LFPs, are continuous time series.Multi-channel Spikes were acquired through high pass filtering the raw data (500-7500Hz), spike detecting and offline denoising, which are the discrete points of potential forms, discrete-time sequence.3. Multi-channel LFPs-Spikes dynamic Spectrum coherence coding1) Discrete points continuedBecause dynamic spectrum coherence coding applied to continuous signals and each channel sequences of Spikes are discrete points, this paper discrete Spikes converts continuous signals through neurons of frequency coding firstly. Select physiology window width 500ms, window step is 125ms, starting from the initial point, calculate average rate per channel each window {r(t)}i, and frequency-coding.2) Pairs of multi-channel LFPs and multi-channel SpikesAfter preprocess of the raw data, based on the physical order of the micro-electrode array, each channel LFPs corresponds to the Spikes, get multi-channel LFPs-Spikes data pairs.3) Dynamic spectrum coherence coding of multi-channel LFPs-SpikesSelect the physiology Windows is 500ms, moving step is 125ms, individually from the initial point of Discrete Prolate Spheroidal Sequences (DPSS) on multi-channel window LFPs-Spikes sequence, calculat for each single channel LFPs-Spikes data on sliding window spectrum coherent value and further cmpute dynamic spectrum coherent values of multichannel LFPs-Spikes. Results1. Working memory training results of rats in Y-mazeThe correct of 17 rats working memory task in a Y-maze increase with the number of training, after 10 days the correct of 17 rats is maintained at more than 80%. Remaining 3 rats who cannot adapt to the Y-maze environment and peanut odor hade been completed excluded.2. Multi-channel raw data of rats in a working memory task in Y-mazeBecause of rats in vivo implantation and record multichannel microelectrode array completed in Y-maze, while recording to LFPs and Spikes of rat 2 over dust-ups tasks:lth 14 channels LFPs and Spikes,2th 15 channel LFPs and Spikes.3th-6th 5,5,6,7channnels. Only LFPs are 5 rats; Only Spikes are 4 rats. Choose to record this thesis to LFPs and Spikes typical of the rats 2 results 10 times, select 10s (working memory before and after the event reference point 5s) as a research object.3. Multiple channel LFPs-Spikes dynamic spectrum coherence coding1) spectrum maps of multichannel LFPs show, the energy are mainly distributed in 0.3-15Hz.2) multi-channel frequency encoding map shows that average discharge rate of multichannel Spikes peak occurred in 2-4s.3)lth rat, repeated 10 times in working memory task, in 2-4s,0.3-15Hz of spectrum coherent values (0.6507±0.012) and other time and frequency spectrum coherent values (0.3182±0.0265), thoug the t test, there are statistical differences (P<0.05).2nd rat repeat 10 times in working memory task, in 2-4s,0.3-15Hz of spectrum coherent values (0.6541±0.0071) and other time and frequency spectrum coherent values (0.3248±0.0061), though the t test, there are statistical differences (P＜0.05).ConclusionsThis thesis study multichannel neural signals of rats (prefrontal cortex) during work and memory in Y-maze, research on multichannel dynamic coherence on working memory of the LFPs-Spikes event code, study results show that:1. Multi-channel in rat prefrontal cortex LFPs-Spikes dynamic spectrum coherece codes memory events effectively:2-4s,0.3-15Hz of spectrum coherent value significantly higher than the other times and frequency (P<0.05), and before the working memory of the time reference point (5s), predicts working memory events.2. Dynamic spectrum coherence of multi-channel LFPs-Spikes from two different modal collaborative perspects encodes working memory events, with supplement of single-mode multi-channel LFPs dynamic spectrum encoding and multi-channel complementary Spikes frequency coding.