Annu Int Conf IEEE Eng Med Biol Soc. 2022 Jul;2022:4040-4043. doi: 10.1109/EMBC48229.2022.9871767.
Imagined speech based brain-computer interface (BCI) is of great interest due to its efficiency and user-friendliness for patients with speech impairment. The aim of this work was to study whether different rhythms of imagined syllables could elicit corresponding frequency components on EEG amplitude spectra. Seventeen participants were recruited to take part in the experiments, and performed a control task and four imagery tasks with the presence of periodic pure tones while their EEG signals were recorded. The four imagery tasks included imagining the syllable' /a/' every time, every two times, and every three times the periodic pure tones occurred, and imagined twice every three times the periodic pure tones occurred. The experimental results analyzed by Fourier transform indicated that neural entrainment to rhythmic speech imagery can be notably reflected on the EEG amplitude spectra. Clinical Relevance- This work manifested that different rhythms of imagined syllables could be identified from EEG amplitude spectra, which may be beneficial to the development of imagined speech based BCIs.
基于想象言语的脑-机接口（BCI）因其对言语障碍患者的高效性和友好性而受到广泛关注。本研究旨在探讨想象不同节律的音节是否能在 EEG 幅度谱上诱发出相应的频率成分。17 名参与者参与了实验，在周期性纯音存在的情况下进行了控制任务和四个想象任务，同时记录他们的 EEG 信号。四个想象任务包括：每次、每两次和每三次周期性纯音出现时想象音节“/a/”，以及每三次周期性纯音出现时想象两次。通过傅里叶变换分析的实验结果表明，对节奏性言语想象的神经同步可以在 EEG 幅度谱上明显反映出来。临床意义- 本工作表明，从 EEG 幅度谱中可以识别出不同节律的想象音节，这可能有助于基于想象言语的 BCI 的发展。
