Neural Decoding of Spontaneous Overt and Intended Speech

Area of Science:

• Neuroscience
• Machine Learning
• Brain-Computer Interface

Background:

• Decoding speech from neural signals is crucial for advancing brain-computer interfaces (BCIs).
• Previous research often relies on cued or imagined speech, limiting real-world applications.

Purpose of the Study:

• To decode intended and overt speech directly from neuromagnetic signals during spontaneous speech tasks.
• To evaluate the effectiveness of machine learning models in classifying speech from neural data without prompts.

Main Methods:

• Magnetoencephalography (MEG) was used to record neural signals from seven healthy adults.
• Participants spontaneously spoke 'yes' or 'no' at a self-paced rate.
• Linear Discriminant Analysis (LDA) and 1D Convolutional Neural Network (1D CNN) were applied for speech decoding.

Main Results:

• The 1D CNN achieved 90.40% accuracy in decoding overt speech, significantly above chance (50%).
• LDA achieved 79.02% accuracy for overt speech decoding.
• The 1D CNN demonstrated 67.19% accuracy in decoding intended speech.

Conclusions:

• Spontaneous overt and intended speech can be decoded directly from neural signals without perceptual interference.
• These findings represent a significant step towards developing spontaneous speech-based BCIs.
• The study highlights the potential of machine learning for real-time speech decoding from neuromagnetic data.