Multidimensional Motor Evoked Potentials (MultiMEP): Digging up buried information from single trials.

Area of Science:

• Neuroscience
• Motor Control
• Brain-Computer Interfaces

Background:

• Transcranial magnetic stimulation (TMS) studies use motor-evoked potentials (MEPs) to infer motor representation states.
• Traditional methods limit analysis to single muscles and simple actions, overlooking intermuscular relationships.
• Univariate analyses of MEPs treat different muscles as independent, missing complex patterns.

Purpose of the Study:

• To overcome limitations in studying covert motor processes using TMS.
• To develop a decoding approach for MEPs, analogous to multivoxel pattern analysis in neuroimaging.
• To investigate richer information encoded in MEPs at single-trial levels.

Main Methods:

• Introduced the novel Multidimensional Motor Evoked Potentials (MultiMEP) approach.
• Applied decoding analysis to MEPs from 24 electrodes during motor imagery of three complex hand actions in 22 participants.
• Conducted cross-classification analysis between motor imagery MEPs and action execution electromyographic patterns.

Main Results:

• Motor imagery of complex hand actions was classified with 74% accuracy using MultiMEP patterns.
• Cross-classification achieved above-chance accuracies: 54% (execution-to-imagery) and 71% (imagery-to-execution).
• Demonstrated that MEPs encode significant information at both single-subject and single-trial levels.

Conclusions:

• MEPs contain richer information than previously recognized, enabling decoding of complex motor processes.
• The MultiMEP decoding approach represents a significant advancement for TMS-based motor control research.
• This method has the potential to shift the paradigm in studying brain-cognition relationships, similar to multivoxel pattern analysis in neuroimaging.