Mysterious brain activity under anesthesia challenges what we know about consciousness
Even under general anesthesia, the human brain can analyze language, recognize different types of words, and predict what comes next in a story, according to a surprising discovery by Baylor researchers. The findings challenge our understanding of consciousness and suggest that the unconscious brain is far more capable than scientists previously realized.
Key points
- Baylor researchers found the hippocampus processes sounds and language during general anesthesia.
- Hippocampal neurons showed plasticity and predicted upcoming words without conscious awareness.
- The study, published in Nature, may aid speech prosthetics and brain-computer interfaces.
Researchers at Baylor College of Medicine have found that the brain retains sophisticated abilities during general anesthesia, including processing language and anticipating upcoming words. The findings, published in the journal Nature, challenge traditional assumptions about consciousness and thinking.
The study examined patients undergoing epilepsy surgery while under general anesthesia, using Neuropixels probes to record electrical signals from hundreds of individual neurons in the hippocampus, a brain structure involved in learning and memory.
What the experiments showed
In one experiment, patients heard repeating tones occasionally interrupted by a different sound. Hippocampal neurons responded differently to the unexpected sounds and became better at distinguishing them over time, suggesting neural plasticity without conscious awareness.
In another experiment, short stories were played to anesthetized patients. Neural activity contained information about word categories such as nouns, verbs, and adjectives, and also carried information that could predict upcoming words in a sentence.
Our findings show that the brain is far more active and capable during unconsciousness than previously thought.
— Dr. Sameer Sheth
Implications and limitations
The researchers note that the results cannot automatically be extended to other anesthetic methods or to sleep and coma. The study focused on the hippocampus, and further work is needed to see if similar abilities exist in other brain regions.
The findings could have practical implications for brain-computer interfaces and speech prosthetics, which help people communicate after neurological injuries. They also highlight similarities between hippocampal prediction and how large language models anticipate text.
