We use cookies in order to improve the quality and usability of the HSE website. More information about the use of cookies is available here, and the regulations on processing personal data can be found here. By continuing to use the site, you hereby confirm that you have been informed of the use of cookies by the HSE website and agree with our rules for processing personal data. You may disable cookies in your browser settings.

  • A
  • A
  • A
  • ABC
  • ABC
  • ABC
  • А
  • А
  • А
  • А
  • А
Regular version of the site

Our article was published in the Journal of Neural Engineering!

The Journal of Neural Engineering published an article "Decoding speech from a small set of spatially separated minimally invasive intracranial EEG electrodes using a compact and interpretable neural network.

Authors: Artur Petrosyan, Alexey Voskoboinikov, Dmitrii Sukhinin, Anna Makarova, Anastasia Skalnaya, Nastasia Arkhipova, Mikhail Sinkin and Alexei Ossadtchi.

Researchers from the Center for Bioelectrical Interfaces of the National Research University Higher School of Economics showed the ability to decode speech using a minimal set of electrodes (6 stereo EEG electrodes on one pin (shaft) and 8 ECoG electrodes on one strip in another patient).

For decoding we used a neural network of our own design, which allows efficient interpretation of its weights, highlighting signal sources contributing to the classification. In comparison with other architectures the solution presented in the article worked equally or better than those described in similar works on speech neural decoding. It was possible to show, that decoding is made at the expense of signals of brain origin, instead of artifacts.
Thus, the possibility of creating a speech prosthesis with a small number of electrodes based on a compact decoder and a small amount of training data was revealed. Once refined, the technology could with high probability provide speech decoding for many severe patients.

The article is available at this link.
Publicly available at biorXiv at the link.