|
Objective: This thesis is based on the Wigner-Ville Distribution (WVD) method, phase synchronization analysis of the rat frontal cortex multi-channel local field potentials (Local Field Potentials, LFPs) in the working memory process, research LFPs with related to memory 0 and 1γ band phase synchronization pattern and its effect on working memory event encoding. And refinement of research in the γ band bandwidth 2Hz each component phase synchronous mode events working memory encoding, support for neural coding mechanism of working memory neural computation. Method: 1. Experimental data: experimental data used in this paper by the Tianjin Medical University Neural Engineering Laboratory. Prefrontal cortex in SD rats before implantation of 16-channel microelectrode array. Used in the body of multi-channel data recording technology, for four rats in the frontal cortex of 50 times training in the Y maze working memory process 16 channels of data. 7s Data Select is sufficient to characterize the whole process of working memory as an object of study. Data preprocessing: low pass filtering the original data recorded to the 16-channel (0.3 to 500Hz), low pass signal obtained is 16 channels LFPS. LFP obtained fitting curve fitting each channel separately, from the the original LFP subtracted fitting trend term, zero mean 16-channel LFPs. 3 LFPs FFT analysis, rats energy is concentrated in the working memory process LFPs band (40-60Hz). 4 different physiological band access to the 16-channel LFPs: application of band-pass filtering method 16-channel LFPs broken down into components and high frequency of various physiological band Y component; 0.3-3 Hz the d wave ,4-12 Hz for θ waves, 12 -15Hz for a wave ,15-30 Hz for the B wave ,30-80 Hz wave γγ ,80-500Hz for the high-frequency Y wave. 5 to extract the instantaneous phase of the 16-channel LFPs: application of the instantaneous phase of the Hilbert transform to extract the 16 channel the LFPs and its various band component. Reference channel selection: Select the highest average discharge frequency channel as the reference channel. Moving window size and mobile steps of selection: select window for 50ms, moving step 12.5ms. 8 multi-channel signal dynamic WVD analysis: (1) of WVD analysis method the initial window starts, to different band component calculated for each window of each channel of LFP phase of the phase sequence of the reference channel nonlinear WVD items one by one; then calculated for each a window in each channel LFP the different band linear phase sequence WVD items were given this band nonlinear and linear dynamic distribution mode WVD study the LFPs synchronous oscillation, although the working memory encoding of events. (2) narrow-band WVD analysis because the Hilbert transform suitable for narrowband signal, in order to verify the validity of the band wider study of the physiological band, the monograph in the coding working memory events critical γ band WVD analysis. The application of wavelet packet analysis scale γ-band (30-80 Hz) wavelet packet decomposition, the decomposition of the signal bandwidth of 2Hz. γ band of 16 to 40, respectively, corresponding to the ninth floor of wavelet packet wavelet packet frequency bands corresponding to the 30-32Hz-78-80Hz. WVD analysis of and bandwidth obtained 2Hz different bands, and wavelet packet component. From the initial window, the calculation of each window channel LFP phase sequence of the different bands the nonlinear the WVD items, the phase sequence of the reference channel and then calculated each window, each channel the LFP different band linear phase sequence WVD entry in the γ band in each 2Hz range of spatial and temporal distribution of the LFP nonlinear and linear WVD synchronous oscillation mode. Results: An analysis of the paper application WVD four SD rats were 10 experiments prefrontal cortex in the working memory of the Y-maze experiment before 16-channel LFPs of θ, γ component phase synchronous mode, and refinement in γ band 2Hz bandwidth of the respective components of the phase synchronization pattern. The main findings are as follows: 1. During working memory events, multi-channel LFPs energy is mainly concentrated in the range of 40-60Hz. Working memory event occurred during the four rats each of 10 experiments in various physiological band reference point before 2s nonlinear WVD value as follows: d-band nonlinear value: 0.1451 ± 0.0231,0.1685 ± 0.0289,0.1972 ± 0.0217,0.1989 ± 0.0328; θ band nonlinear value: .1771 ± 0.0134,0.1834 ± 0.0122,0.1619 ± 0.0105,0.1480 ± 0.0139; the a bands nonlinear value: 0.1560 ± 0.0654,0.1774 ± 0.0736,0.2013 ± 0.0105,0.1881 ± 0.0605; β band nonlinear value: 0.1555 ± 0.0644,0.1814 ± 0.0559,0.2020 ± 0.0228,0.1585 ± 0.0393; gamma band nonlinear value: 0.1742 ± 0.0114,0.1364 ± 0.0126,0.1671 ± 0.0423,0.1618 ± 0.0372. T-test, the rat prefrontal cortex multi-channel LFPs θ and γ band nonlinear WVD items the the apparent synchronous oscillation (P lt; 0.05). D, a, β band nonlinear WVD items without synchronous oscillation (P gt; 0.05). Working memory event occurred during the period, four rats each of 10 experiments in various physiological band reference point before 2s linear WVD value as follows: d-band linear value: 0.3769 ± 0.0314,0.3654 ± 0.0270,0.3723 ± 0.0361 ± 0.0284 .3438; theta band linear value: 0.3449 ± 0.0121,0.3505 ± 0.0104,0.4076 ± 0.0232,0.3947 ± 0.0106; a band linear: .3660 ± 0.0322,0.3565 ± 0.0313,0.3682 ± 0.0418,0.3546 ± 0.0230 ; β-band linear values: .3665 ± 0.0296,0.3525 ± 0.0204,0.3675 ± 0.0415,0.3842 ± 0.0331; gamma-band linear value: 0.3478 ± 0.0619,0.3975 ± 0.0368,0.4024 ± 0.0376,0.3809 ± 0.0421. By t test, the rat frontal cortex multi-channel LFPs in the θ band linear WVD items obvious synchronous oscillation (P lt; 0.05), d, a, β, γ-band linear WVD no synchronous oscillation ( P gt; 0.05). Working memory events during the four rats each experiment before the 40-42Hz-58-60Hz reference point 2s within the nonlinear WVD value of as follows :40-42Hz nonlinear values: 0.1778 ± 0.0121,0.1987 ± 0.0017,0.2301 ± 0.0193,0.2282 ± 0.0164; 42-44Hz nonlinear value: .1782 ± 0.0392,0.1941 ± 0.0433,0.2235 ± 0.0587,0.2154 ± 0.0381; 44-46Hz nonlinear value: 0.2428 ± 0.0282,0.2266 ± 0.0031 , 0.2901 ± 0.0174,0.3227 ± 0.0137; 46-48Hz nonlinear value: 0.1792 ± 0.0091,0.1960 ± 0.0150,0.2220 ± 0.0162,0.2071 ± 0.0013; 48-50Hz nonlinear value: 0.1795 ± 0.0310,0.1741 ± 0.0372,0.2162 ± 0.0407,0.2045 ± 0.0341; 50-52Hz nonlinear value: 0.1528 ± 0.0293,0.1706 ± 0.0396,0.1953 ± 0.0462,0.1870 ± 0.0334; 52-54Hz nonlinear value: 0.1489 ± 0.0292,0.1684 ± 0.0181,0.1990 ± 0.0227 , 0.1959 ± 0.0235; 54-56Hz nonlinear value: 0.1763 ± 0.0081,0.2011 ± 0.0133,0.2207 ± 0.0125,0.2162 ± 0.0120; 56-58Hz nonlinear value: 0.1608 ± 0.0113,0.2070 ± 0.0142,0.2316 ± 0.0108,0.2231 ± 0.0054; 58-60Hz nonlinear: 0.1725 ± 0.0169,0.2040 ± 0.0128,0.2276 ± 0.0115,0.2242 ± 0.0137. By t test, the rat frontal cortex multi-channel LFPs occurred in the 40-42Hz ,44-46Hz ,46-48Hz ,54-56Hz ,56-58Hz ,58-60Hz nonlinear WVD items in apparent synchronous oscillation (P lt ; 0.05). 42-44Hz ,48-50Hz ,50-52Hz ,52-54Hz nonlinear WVD items synchronous oscillation (P gt; 0.05) did not happen. During working memory events, four rats each experiment before the 40-42Hz ~ 58-60Hz reference point 2s linear WVD values ??as follows :40-42Hz linear value: 0.3442 ± 0.0685,0.3352 ± 0.0623,0.3394 ± 0.0664,0.3145 ± 0.0661; 42-44Hz linear value: 0.3438 ± 0.0698,0.3398 ± 0.0512,0.3460 ± 0.0494,0.3273 ± 0.0661; 44-46Hz linear value: 0.2792 ± 0.0235,0.3073 ± 0.0201,0.2794 ± 0.0170,0.2200 ± 0.0194; 46-48Hz linear values: .3438 ± 0.0789,0.3379 ± 0.0553,0.3475 ± 0.0526,0.3256 ± 0.0461; 48-50Hz linear value: 0.3425 ± 0.0498,0.3598 ± 0.0523,0.3533 ± 0.0416,0.3382 ± 0.0615; 50-52Hz linear values: .3692 ± 0.0406,0.3633 ± 0.0378,0.3742 ± 0.0270,0.3557 ± 0.0217; 52-54Hz linear value: 0.3731 ± 0.0282,0.3655 ± 0.0312,0.3705 ± 0.0195,0.3468 ± 0.0267; 54 - 56Hz linear value: 0.3457 ± 0.0790,0.3328 ± 0.0596,0.3488 ± 0.0516,0.3265 ± 0.0474; 56-58Hz linear value: 0.3612 ± 0.0522,0.3269 ± 0.0796,0.3379 ± 0.0682,0.3196 ± 0.0556; 58-60Hz linear value : 0.3495 ± 0.0673,0.3299 ± 0.0784,0.3419 ± 0.0589,0.3185 ± 0.0563. T-test, the rat prefrontal cortex multi-channel LFPs linear WVD items in the 44-46Hz obvious synchronous oscillation (P lt; 0.05). 40-42Hz ,42-44Hz ,46-48Hz ,48-50Hz ,50-52Hz ,52-54Hz ,54-56Hz ,56-58Hz ,58-60Hz linear WVD did not happen synchronous oscillation (P gt; 0.05 ). Conclusion: in the process of working memory in the Y-maze in rats, the prefrontal cortex multi channel LFPs concentrated in the 40-60Hz energy, is closely related to the band and working memory. In the process of working memory in the Y-maze in rats applications WVD analysis LFPs found the the nonlinear WVD items of θ and γ band synchronous oscillation apparent effective coding working memory events: d, a β-band nonlinear WVD The entries are not synchronized oscillation can not be effective coding working memory events. The Application the WVD Analysis LFPs find 0 band linear WVD synchronous oscillation apparent WVD γ-band linear synchronous oscillation. Indicating the validity of the analysis WVD. 3 (1) in the working memory of the Y-maze in rats event process applications WVD analysis LFPs, found in the γ band 40-42Hz ,44-46Hz ,46-48Hz ,54-56Hz ,56-58Hz ,58-60Hz nonlinear WVD synchronous oscillation significantly effective coding working memory events ,42-44H ,48-50Hz, 50-52Hz ,52-54Hz nonlinear the WVD items not synchronized oscillation occurs, can not be effective coding work memory of the event. (2) in the working memory of the Y-maze in rats event procedure, the application WVD analysis LFPs, the entire the γ band can not reflect LFPs in the gamma-band synchronization coding, but the gamma band 44-46Hz synchronous oscillation, can be encoded in working memory events .
|