Introduction
Understanding the cortical representation of movement is central to the study of human neurophysiology. While it is recognized that both efferent motor commands and afferent sensory feedback contribute to motor control [1][2], their distinct and combined roles have yet to be fully studied [3]. Our research examines this relationship by analyzing neural signals during hand grip behaviours, employing Electroencephalogram (EEG) and functional near-infrared spectroscopy (fNIRS) for functional brain imaging. Furthermore, we utilized Functional Electrical Stimulation (FES) to stimulate muscle contraction in the absence of voluntary motor commands from the brain, solely incorporating sensory feedback of movement [4]. This study aims to contribute to the broader discussion on the neurophysiological bases of human movement, promoting a deeper understanding of motor control mechanics.
Methods
We designed 4 distinct task conditions to acquire different combinations of sensory feedback and motor commands in hand grip behaviors (as shown in Table 1) and conducted statistical analyses to investigate the interplay between them. We simultaneously recorded the EEG and fNIRS signals of eight right-dominant participants during bimanual hand grip tasks under the 4 conditions. The force was normalized based on each individual’s maximum voluntary grip power during Maximum Voluntary Contraction (MVC) of muscles, which was measured prior to the experiments.
The participants were guided via a GUI throughout the tasks as shown in Figure 1. Our unique sensor configuration enabled an efficient fusion of EEG and fNIRS data. Post-data acquisition, Pearson correlation coefficients were computed to ascertain multimodality correlations between EEG, oxygenated hemoglobin (HbO), deoxygenated hemoglobin (HbR), and grip force, and task-specific EEG signal correlations.
Results
The EEG mu-band spectrograms from each session as shown in Figure 2 exhibited power decreases along with the periodic grip or MI, known as mu-band event-related desynchronization, an indication of motor cortex activation. The congruence of EEG signal changes with hand grip behavior in both time and frequency domains affirms the capture of motor-related brain activity.
The multimodal correlation maps, averaged across subjects, displayed in Figure 3 showed opposite correlations between EEG and force were observed in distinct EEG hemispheres, with strong correlations in ipsilateral EEG and force, for tasks with sensory feedback. A robust anti-correlation existed between HbO and HbR in tasks incorporating sensory feedback, signifying brain activation. Such correlations underscore valid brain activation dynamics which affirmed the data set's integrity.
Regarding EEG task correlations, the Passive-FES-grip task exhibited minimal correlation with the Pure-Motor-imagery task, implying limited overlap between sensory feedback and motor command as shown in Figure 4. While all tasks correlated positively with natural voluntary grip, the pure-MI task displayed a higher correlation than the passive-FES task (p=0.0222), suggesting shared brain components but the dominance of motor commands.
Conclusion
Our findings illuminate the distinct representations of motor command and sensory feedback, emphasizing the former's alignment with natural voluntary signals. The use of FES to trigger sensory feedback, independent of brain activity, underlines the potential to enrich sensory feedback understanding. In conclusion, our research challenges existing paradigms and paves the way for a refined understanding of the neurophysiological bases of human movement.
