Comparing effects in spike-triggered averages of rectified EMG across different behaviors

Abstract

Effects in spike-triggered averages (SpikeTAs) of rectified electromyographic activity (EMG) compiled for the same neuron-muscle pair during various behaviors often appear different. Do these differences represent significant changes in the effect of the neuron on the muscle activity? Quantitative comparison of such differences has been limited by two methodological problems, which we address here. First, although the linear baseline trend of many SpikeTAs can be adjusted with ramp subtraction, the curvilinear baseline trend of other SpikeTAs can not. To address this problem, we estimated baseline trends using a form of moving average. Artificial triggers were created in 1 ms increments from 40 ms before to 40 ms after each spike used to compile the SpikeTA. These 81 triggers were used to compile another average of rectified EMG, which we call a single-spike increment-shifted average (single-spike ISA). Single-spike ISAs were averaged to produce an overall ISA, which captured slow trends in the baseline EMG while distributing any spike-locked features evenly throughout the 80 ms analysis window. The overall ISA then was subtracted from the initial SpikeTA, removing any slow baseline trends for more accurate measurement of SpikeTA effects. Second, the measured amplitude and temporal characteristics of SpikeTA effects produced by the same neuron-muscle pair may vary during different behaviors. But whether or not such variation is significant has been difficult to ascertain. We therefore applied a multiple fragment approach to permit statistical comparison of the measured features of SpikeTA effects for the same neuron-muscle pair during different behavioral epochs. Spike trains recorded in each task were divided into non-overlapping fragments of 100 spikes each, and a separate, ISA-corrected, SpikeTA was compiled for each fragment. Measurements made on these fragment SpikeTAs then were used as test statistics for comparison of peak percent increase, mean percent increase, peak width at half maximum, onset latency, and offset latency. The average of each test statistic measured from the fragment SpikeTAs was well correlated with the single measurement made on the overall SpikeTA. The multiple fragment approach provides a sensitive means of identifying significant changes in SpikeTA effects.

Figure 1

Measurements made on SpikeTAs. A shows an 80 ms SpikeTA of rectified EMG from FDPu compiled from 24,470 EMG-filtered spikes discharged by neuron e0030 during an RPD epoch. Three horizontal lines indicate the levels of the mean ± 2 standard deviations (SDs) of the SpikeTA during the baseline sample period. The points at which the SpikeTA crosses the ± 2 SDs line before and after the peak define the Onset and Offset, of the SpikeTA effect respectively. A separate horizontal line segment indicates the peak width at half maximum (PWHM) of the SpikeTA effect. Three vertical line segments indicate the time of the beginning of the spike (action potential), the time of the trigger pulses used to align data for averaging, and the Peak Amplitude above the baseline sample period mean. Algorithms for making these and other measurements are described in the Methods. B shows an 80 ms fragment SpikeTA compiled from only 100 of the spikes that contributed to the overall SpikeTA in A. As described in the Results, the same algorithms used to measure features of the overall SpikeTA were used to measure features of the fragment SpikeTA. The ordinates in A and B have been scaled to fill the same vertical height from minimum to maximum of the SpikeTA waveforms. For both A and B, units of average rectified EMG ( $\overline{\text{rEMG}}$) are in Volts after amplification 10,000x.

Figure 2

Adjustment of baseline trends with Increment Shifted Averages. A-F show SpikeTAs from 6 different neuron-muscle pairs with various baseline trends: A – flat; B – rising ramp; C – falling ramp; D – convex up; E – convex down; F – inflection. The left column shows the initial SpikeTA for each neuron-muscle pair. To reveal the baseline trends more clearly, in this column 160 ms are shown, although analyses typically are performed on SpikeTAs 80 ms in duration. The middle column shows the increment shifted average (ISA) estimate of the baseline trend during the 80 ms period used for analysis. The right column shows the 80 ms ISA-adjusted SpikeTA. Note that in these adjusted SpikeTAs the baseline trend has been effectively flattened, without reducing the stochastic noise of the SpikeTA or altering the height of the SpikeTA above 0. For each neuron-muscle pair A – F, the same ordinate scale at left ( $\overline{\text{rEMG}}$ in Volts after amplification) applies to all three averages in the row: initial SpikeTA, ISA baseline, and adjusted SpikeTA. A vertical hairline has been drawn at the trigger time for each average. Time scales beneath each average show major tick marks at the trigger time (0 ms), and at −30 ms and +50 ms (the 80 ms analysis window), with minor tick marks every 5 ms.

Figure 3

Increment shifted averaging to estimate the baseline trend. A. A short segment of a microelectrode recording shows three large spikes discharged by a neuron. EMG recorded simultaneously from EDC is shown below. B. Triggers from each of the three neuron spikes in A are are shown above, and corresponding segments of full-wave rectified EMG are shown below. C. For each of the three SpikeTA triggers in B, 81 artificial triggers were generated at increments of 1 ms, from 40 ms before to 40 ms after the SpikeTA trigger. These 81 artificial triggers were used to define corresponding segments of rectified EMG which then were aligned and averaged to compile the single-spike increment shifted averages (ISAs) shown below. The 1kHz periodicity of the single-spike ISAs reflects the 1 ms interval between artificial triggers. Nevertheless, the noise level of the single-spike ISAs shown in C is substantially lower than the variability of the full-wave rectified EMG shown in B (note vertical scales at right). All data in A, B and C are aligned in time (scale in C). D. The initial SpikeTA for this neuron-muscle pair compiled from 6,014 segments of rectified EMG like the three illustrated in B showed a rising curvilinear baseline (left). This baseline was captured by the overall ISA (center), compiled from 6,014 corresponding single-spike ISAs, like the three shown in C. Subtraction of the overall ISA from the initial SpikeTA produced the adjusted SpikeTA (right) with a relatively flattened baseline but preserved stochastic noise level.

Figure 4

A. SpikeTAs compiled for the same neuron-muscle pair during two different behavioral epochs are shown on the same vertical scale (Volts after amplification 10,000x). Left, Epoch 1 – squeeze task; Right, Epoch 5 – RPD of neuron e0030 versus FDPu. Both SpikeTAs are 80 ms in duration. B. Fifteen sequential fragment spike-triggered averages (fSpikeTAs), each compiled from 100 sequential spikes, are shown from Epoch 1 (left) and from Epoch 5 (right). Each fSpikeTA is scaled to fill the same vertical height from minimum to maximum, and is 80 ms in duration. C. The test statistic, d, is plotted as a function of time for each of the fragments from Epoch 1 (left) and Epoch 5 (right). The time at which each d value is plotted is the time at which that fragment ended. The time scale of 1000 s (left) applies to both epochs. Values of d for the 15 fSpikeTAs shown in B are indicated with horizontal curly brackets.

Figure 5

A shows SpikeTAs for neuron-muscle pair e0030-FDPu compiled for each of nine successive behavioral epochs: 1) Squeeze task; 2) RPD of neuron e0030 versus FCR; 3) e0030 v FDPu; 4) e0030 v FCR; 5) e0030 v FDPu; 6) e0030 v APL; 7) e0030 v ECRB; 8) e0030 v FCU; 9) e0030 v FDPu. To facilitate visual comparison of the peaks, here these nine SpikeTAs have been scaled individually to match the baseline level of EMG relative to zero. The abscissa time scale is 5000 s overall (scale at bottom of Figure), with the points at which behavioral epochs changed indicated with open squares. B shows test statistics fPPI, fPWHM, fOnset, fOffset, and fMPI for each fSpikeTA plotted as a function of time. Each value is plotted at the time at which the last spike of that fragment occurred. Abcissas again are 5000 s long, with tics every 1000 s, and open squares at the borders between behavioral epochs.

Figure 6

Correlations between test statistics and measures of overall SpikeTA effects. Scatterplots illustrate the correlations between the average of fragment test statistics (ordinate: fPPI, fMPI, fPWHM, fOnset and fOffset) and measurements of amplitude (A) and temporal (B) features of SpikeTA effects (abscissa: PPI, MPI, PWHM, Offset and Onset, respectively) made from the overall SpikeTAs during nine different behavioral epochs for neuron-muscle pair e0030-FDPu.

Figure 7

Effect of the number of spikes per fragment and of the number of fragments on P-values. A. SpikeTAs from nine behavioral epochs are shown for a neuron-muscle pair selected for the presence of a clear SpikeTA effect in several epochs (2, 3, 4, 5, 6, and 8), no effect in other epochs (1 and 7), and a borderline effect in one epoch (9). For the SpikeTA in epoch 9, a questionable peak might be present (*), but features of similar size are present at other times (^). All nine SpikeTAs are scaled to fill the same vertical height from minimum to maximum, and are 80 ms in duration. The P-value obtained testing the null hypothesis that no SpikeTA effect (peak) is present is shown as a function of the number of spikes per fragment in B, and as a function of the number of fragments in C, for each of the nine epochs (test-statistic d, Wilcoxon signed rank test). A horizontal line has been drawn in B and C at P = 0.005 (0.05 after Bonferonni correction for the 9 tests on 9 epochs), and the legend in B applies to C as well. The P-value obtained testing the null hypotheses that measured parameters of the SpikeTA effects (PPI, MPI, PWHM, Onset and Offset) did not change across behavioral epochs is shown for each parameter (test statistics fPPI, fMPI, fPWHM, fOnset and fOffset, respectively; Kruskal-Wallis tests) as a function of the number of spikes per fragment in D and as a function of the average number of fragments across the nine epochs in E. A horizontal line has been drawn in D and E at P = 0.01 (0.05 after Bonferonni correction for 5 tests). A floor of log(P) = −20 was used to permit better visualization of these results for parameters other than PPI. The legend in D applies to E as well Note that in B, C, D and E both abscissa and ordinate are logarithmic scales.