Anesthetized Brains Still Decode Language, Predict
Get the Health newsletter
Daily health & science — research, biotech, public health, the studies worth knowing. Free.
- Researchers at Baylor College of Medicine found that patients under general anesthesia could still process language at a sophisticated level, distinguishing nouns, verbs, and adjectives while listening to short stories.
- Dr. Sameer Sheth and colleagues used Neuropixels probes to record hundreds of individual neurons in the hippocampus of epilepsy surgery patients — the first time the technology had been deployed in the hippocampus for this type of research.
- Hippocampal neurons not only detected unexpected tones embedded in repeating sounds but also grew better at recognizing them over time, indicating that neural plasticity continues during anesthesia.
- Neural signals in the hippocampus could be used to predict upcoming words before they were spoken, a form of predictive coding the researchers said they had previously associated only with waking attention.
- Consciousness may arise from communication across multiple brain regions rather than from activity within any single area like the hippocampus, the Nature paper argues.
- The hippocampus's predictive behavior mirrored how large language models generate text by anticipating the next word, the researchers noted, hinting at shared principles between biological and artificial intelligence.
- Dr. Katlowitz raised the prospect of deploying these neural signals to power speech prosthetics for people who lost the ability to speak due to stroke or injury.
Why it matters: For decades, scientists treated consciousness as a prerequisite for sophisticated cognition. Baylor's Nature study shows the hippocampus runs predictive language processing in fully anesthetized patients, and the researchers explicitly raised the prospect of using these hippocampal signals to build speech prosthetics for people who lost speech to stroke or injury.
Ask SkimNews
