BackgroundBrain-machine interfaces (BMIs) involving electrodes implanted into the human cerebral cortex have recently been developed in an attempt to restore function to profoundly paralyzed individuals. Current BMIs for restoring communication can provide important capabilities via a typing process, but unfortunately they are only capable of slow communication rates. In the current study we use a novel approach to speech restoration in which we decode continuous auditory parameters for a real-time speech synthesizer from neuronal activity in motor cortex during attempted speech.Methodology/Principal FindingsNeural signals recorded by a Neurotrophic Electrode implanted in a speech-related region of the left precentral gyrus of a human volunteer suffering from locked-in syndrome, characterized by near-total paralysis with spared cognition, were transmitted wirelessly across the scalp and used to drive a speech synthesizer. A Kalman filter-based decoder translated the neural signals generated during attempted speech into continuous parameters for controlling a synthesizer that provided immediate (within 50 ms) auditory feedback of the decoded sound. Accuracy of the volunteer's vowel productions with the synthesizer improved quickly with practice, with a 25% improvement in average hit rate (from 45% to 70%) and 46% decrease in average endpoint error from the first to the last block of a three-vowel task.Conclusions/SignificanceOur results support the feasibility of neural prostheses that may have the potential to provide near-conversational synthetic speech output for individuals with severely impaired speech motor control. They also provide an initial glimpse into the functional properties of neurons in speech motor cortical areas.

We investigated the effects of semantic priming on initial encoding of briefly presented pictures of objects and scenes. Pictures in four experiments were presented for varying durations and were followed immediately by a mask. In Experiments l and 2, pictures of simple objects were either preceded or not preceded by the object's category name (e.g., dog). In Experiment 1 we measured immediate object identification; in Experiment 2 we measured delayed old/new recognition in which targets and distractors were from the same categories. In Experiment 3 naturalistic scenes were either preceded or not preceded by the scene's category name (e.g., supermarket). We measured delayed recognition in which targets and distractors were described by the same category names. In Experiments 1-3, performance was better for primed than for unprimed pictures. Experiment 4 was similar to Experiment 2 in that we measured delayed recognition for simple objects. As in Experiments 1-3, a prime that preceded the object improved subsequent memory performance for the object. However, a prime that followed the object did not affect subsequent performance. Together, these results imply that priming leads to more efficient information acquisition. We offer a picture-processing model that accounts for these results. The model's central assumption is that knowledge of a picture's category (gist) increases the rate at which visual information is acquired from the picture.It is well known that preceding a target picture by a semantically related priming stimulus can speed its identification. The typical explanation is that primes influence relatively late cognitive processes--that is, processes for which input is perceptual information acquired from the target picture. For instance, primes might reduce the amount of stimulus information required to gain access to entries in lexical memory or meaning representations in semantic memory (e.g., Ratcliff & McKoon, 1988).However, primes may also influence early perceptual processing; that is, they may affect acquisition of initial, perceptualThe research was supported by a grant from the National Institute of Mental Health to Geoffrey R. Loftus. Experiment 1 is part of a doctoral dissertation submitted to the University of Washington by Mark Tippens Reinitz and was presented by Reinitz and Loftus at the 1986 Psychonomic Society meeting in New Orleans. Experiments 2-4 were presented by Reinitz, Eve Wright, and Loftns at the 1988 Psychonomic Society meetings in Chicago.Our heartfelt thanks go to Rose Ahem, Janine Duncan, Paul Gerhig, and Susan Pierce for running massive numbers of subjects and for their insights, ideas, criticisms, and comments during all phases of the research. A large number of people provided helpful comments on earlier versions of the manuscript, including Tom Carr, Buz Hunt, Beth Loffus, Steven Lupker, Michael Masson, John Miyamoto, John Palmer, Jonathan Schooler, Gay Snodgrass, George Wolford, and an anonymous reviewer. Mark Tippens Reinitz thanks Southeastern Louisiana University...

No abstract