Decoding Speech Using Brain-Computer Interfaces (dSPEECH)

  • Research type

    Research Study

  • Full title

    Decoding Speech Using Brain-Computer Interfaces based on Intracranial Brain Signals (dSPEECH)

  • IRAS ID

    321659

  • Contact name

    Neil Barua

  • Contact email

    Neil.Barua@nbt.nhs.uk

  • Sponsor organisation

    North Bristol NHS Trust

  • Duration of Study in the UK

    1 years, 9 months, 0 days

  • Research summary

    Many patients with late-stage neurodegenerative diseases such as motor neurone disease (MND) have impaired speech, or even complete loss of speech. Assistive communication devices for these patients do exist, but they are full of limitations. For instance, text-to-speech devices are slow, and sometimes physically demanding for the patient to use.

    For these patients, using such devices can be frustrating because they do not allow for the full, fluent and natural communication that the patient had before development of their disease. As such, many patients are reluctant to use these assistive technologies to communicate with friends and loved ones. This leads to further frustration, and can even lead to the patient not wanting to speak anymore.

    However, in recent years, there much progress has been made in the field of brain-computer interfaces (BCIs). BCIs are devices that read the brain signal, and translate this into a command for an external device. This external device may be, for instance, a cursor on a computer screen: the user thinks for the cursor to go left, and it does. No movement is required from the user: only thought.

    The same is achievable for speech: the user thinks of a word or phrase, and the external device can speak that word or phrase on behalf of the user.

    This study (dSPEECH) aims to further develop BCIs for speech. You may have heard of text-to-speech technologies: dSPEECH aims to create brain-to-speech technologies, where only thought is required to give a voice back to patients who have lost theirs.

  • REC name

    London - Bromley Research Ethics Committee

  • REC reference

    22/PR/1435

  • Date of REC Opinion

    8 Nov 2022

  • REC opinion

    Further Information Favourable Opinion