Modulation of post-movement beta rebound by contraction force and rate of force development

Abstract

Movement induced modulation of the beta rhythm is one of the most robust neural oscillatory phenomena in the brain. In the preparation and execution phases of movement, a loss in beta amplitude is observed [movement related beta decrease (MRBD)]. This is followed by a rebound above baseline on movement cessation [post movement beta rebound (PMBR)]. These effects have been measured widely, and recent work suggests that they may have significant importance. Specifically, they have potential to form the basis of biomarkers for disease, and have been used in neuroscience applications ranging from brain computer interfaces to markers of neural plasticity. However, despite the robust nature of both MRBD and PMBR, the phenomena themselves are poorly understood. In this study, we characterise MRBD and PMBR during a carefully controlled isometric wrist flexion paradigm, isolating two fundamental movement parameters; force output, and the rate of force development (RFD). Our results show that neither altered force output nor RFD has a significant effect on MRBD. In contrast, PMBR was altered by both parameters. Higher force output results in greater PMBR amplitude, and greater RFD results in a PMBR which is higher in amplitude and shorter in duration. These findings demonstrate that careful control of movement parameters can systematically change PMBR. Further, for temporally protracted movements, the PMBR can be over 7 s in duration. This means accurate control of movement and judicious selection of paradigm parameters are critical in future clinical and basic neuroscientific studies of sensorimotor beta oscillations. Hum Brain Mapp 37:2493-2511, 2016. © 2016 Wiley Periodicals, Inc.

Keywords: MEG; event-related desynchronization; event-related synchronization; magnetoencephalography; movement-related beta decrease; neural oscillations; post movement beta rebound; sensorimotor cortex.

Figure 1

(A) A photograph of the isometric wrist‐flexion dynamometer. (B/C) Each target force profile (black) with single examples of real‐time visual feedback showing contraction force overlaid (red). (B) Target profiles for the constant‐force contractions at 5%, 15%, 35% and 60% maximal voluntary force. (C) Target profiles for the ramp contractions, with rates of force development of 86.7%, 28.9% and 10.4% maximum voluntary force output per second. (D) A schematic diagram of the constant‐force contractions experiment. (E) A schematic diagram of the ramp contractions experiment. [Color figure can be viewed in the online issue, which is available at http://wileyonlinelibrary.com.]

Figure 2

Schematic diagram showing the MEG data analysis pipeline.

Figure 3

Results of the constant‐force experiment. (A/E) Spatial signatures of MRBD (A) and PMBR (E) in a single subject. (B/F) Time‐frequency spectrograms extracted from locations of interest at the peak MRBD (B) and PMBR (F); upper to lower panels represent (prescribed) 5%MVF, 15%MVF, 35%MVF and 60%MVF contractions. Note that for MRBD, time zero indicates contraction onset; for PMBR, time zero indicates contraction offset. (C/G) Mean MRBD during the contraction (C) and total PMBR integral over the 10 s post contraction period (G) plotted against force. (D/H) The null distribution (blue) with the measured MRBD (D) and PMBR (H) gradient from real data overlaid in red. Note significant (p_corrected < 0.05) modulation of PMBR with force output was observed. [Color figure can be viewed in the online issue, which is available at http://wileyonlinelibrary.com.]

Figure 4

Results of the ramp experiment. (A/E) Spatial signatures of MRBD (A) and PMBR (E) in a single subject. (B/F) Time‐frequency spectrograms extracted from locations of interest at the peak MRBD (B) and PMBR (F); upper to lower panels represent (prescribed) 10.4%MVF·s⁻¹, 28.9%MVF·s⁻¹, and 86.7%MVF·s⁻¹ contractions. Note that for MRBD, time zero indicates contraction onset; for PMBR, time zero indicates contraction offset. (C/G) Mean MRBD during the contraction (C) and total PMBR integral over the 10 s post contraction period (G) plotted against RFD. (D/H) The null distribution (blue) with the measured MRBD (D) and PMBR (H) gradient from real data overlaid in red. Note no significant modulation of either MRBD or PMBR with RFD. [Color figure can be viewed in the online issue, which is available at http://wileyonlinelibrary.com.]

Figure 5

Measurement of PMBR amplitude and duration during the constant‐force experiment. (A) Shows the average beta band envelope time‐courses (blue) for each of the four force outputs; time t = 0 denotes contraction offset. The black lines show the estimated duration and mean amplitude of the PMBR. (B) Duration of PMBR plotted against force output (upper panel) and the measured gradient (PMBR duration against force) (red line) alongside the null distribution in blue (lower panel). (C) Mean amplitude of PMBR plotted against force output (upper panel) and the measured gradient (PMBR amplitude against force) (red line) alongside the null distribution in blue (lower panel). Note that whilst amplitude of PMBR increases significantly with force output, duration fails to reach significance following FDR correction. [Color figure can be viewed in the online issue, which is available at http://wileyonlinelibrary.com.]

Figure 6

Measurement of PMBR amplitude and duration during the ramp experiment. (A) Shows the average beta band envelope time‐courses (blue) for each of the three RFDs; time t = 0 denotes contraction offset. The black lines show the estimated duration and mean amplitude of the PMBR. (B) Duration of PMBR plotted against RFD (upper panel) and the measured gradient (PMBR duration against RFD) (red line) alongside the null distribution in blue (lower panel). (C) Mean amplitude of PMBR plotted against RFD (upper panel) and the measured gradient (PMBR amplitude against force) (red line) alongside the null distribution in blue (lower panel). Note that as RFD is increased (and the duration of the ramp contraction decreased) the duration of the PMBR is significantly reduced, and PMBR amplitude is significantly increased. [Color figure can be viewed in the online issue, which is available at http://wileyonlinelibrary.com.]

Figure 7

(A/B) Peak locations of the MRBD (blue) and PMBR (red) for each individual subject (top row) and the group average (bottom row), in both the constant‐force (A) and ramp (B) experiments. [Color figure can be viewed in the online issue, which is available at http://wileyonlinelibrary.com.]

Figure A1

Correlation between subject accuracy and MRBD/PMBR: (A) MRBD vs mean error in force; constant‐force experiment. (B) MRBD vs mean error in force; RFD experiment. (C) PMBR vs mean error in force; constant‐force experiment. (D) PMBR vs mean error in force; RFD experiment.