A neural architecture for imagined and overt speech motor dynamics.

High-resolution electrocorticography reveals distinct yet interconnected neural dynamics for speech imagery and articulation, with median decoding accuracy over 80%.

• Why it matters: Understanding the neural basis of speech imagery is crucial for advancing brain-computer interfaces and unraveling human cognitive processes, yet its neural mechanisms are not well characterized.
• What they did: The study used electrocorticography to analyze articulatory kinematic trajectories in frontoparietal and sensorimotor regions during speech imagery and overt speech, employing linear modeling and a generalized decoding framework.
• The result: Findings demonstrate shared supramodal planning and modality-specific representations organized somatotopically, enabling high-accuracy prediction of speech imagery and informing future neural interface development.