Plasticity and language in the anaesthetized human hippocampus

Area of Science:

• Neuroscience
• Cognitive Science
• Anesthesiology

Background:

• Consciousness is central to cognition, but its necessity for higher-order pattern recognition is debated.
• Understanding cognitive processing during unconsciousness is crucial for anesthesiology and neuroscience.

Purpose of the Study:

• To investigate whether complex sensory processing, including pattern recognition and prediction, persists in the human hippocampus during unconsciousness induced by general anesthesia.
• To explore the neural mechanisms underlying such processing in the absence of conscious awareness.

Main Methods:

• Recorded high-density neural activity (single-unit and local-field potentials) in the human hippocampus using Neuropixels microelectrodes.
• Presented auditory stimuli (tones and language) to anesthetized patients.
• Utilized a recurrent neural network model to simulate learning and oddball representation.

Main Results:

• Hippocampal neurons and local oscillations detected oddball tones, with effect size increasing over time, indicating representational plasticity.
• Neural activity encoded semantic and grammatical features of natural speech.
• Hippocampal processing predicted semantic information of upcoming words.
• A neural network model demonstrated emergent learning and oddball representation properties.

Conclusions:

• Complex sensory processing, including auditory oddball discrimination and language comprehension, occurs in the human hippocampus even during general anesthesia-induced unconsciousness.
• These findings suggest that higher-order cognitive functions can persist independently of conscious awareness.
• The hippocampus plays a role in processing complex stimuli irrespective of conscious state, challenging previous assumptions.