BCIwiki (bciwiki.com) — A closed-loop system driven by high-resolution intracranial EEG identified which talker a listener was attending to and amplified that voice in real time, improving speech intelligibility in multi-talker settings, researchers reported in Nature Neuroscience on May 12, 2026. The authors are affiliated with the Department of Electrical Engineering at Columbia University, the Mortimer B. Zuckerman Mind Brain Behavior Institute, Hofstra Northwell School of Medicine and The Feinstein Institutes for Medical Research, among others.
Understanding speech in noise is hard for many people, and current hearing aids often fall short because they amplify all sounds rather than the talker of interest, the researchers wrote. Auditory attention decoding (AAD) uses a listener's brain signals to identify and boost the attended speaker, but whether it could deliver real-time perceptual benefits had been unclear.
The team recorded from patients undergoing neurosurgical procedures and reached the decoding fidelity needed to dynamically amplify the attended talker. Across multiple experiments, the system improved speech intelligibility, reduced listening effort and was consistently preferred by subjects. It also tracked both instructed and self-initiated attention shifts.
The authors say the work provides direct evidence that a real-time, brain-controlled hearing system can enhance perception and sets a performance benchmark for future auditory brain-computer interfaces (BCIs). The approach relied on intracranial electrodes in surgical patients. The paper's PMID is 42115433 and its DOI is 10.1038/s41593-026-02281-5.