Self-paced movement intention detection from human brain signals: Invasive and non-invasive EEG

Abstract

Abstract-Neural signatures of humans' movement intention can be exploited by future neuroprosthesis. We propose a method for detecting self-paced upper limb movement intention from brain signals acquired with both invasive and noninvasive methods. In the first study with scalp electroencephalograph (EEG) signals from healthy controls, we report single trial detection of movement intention using movementrelated potentials (MRPs) in a frequency range between 0.1 to 1 Hz. Movement intention can be detected above chance level (p<0.05) on average 460 ms before the movement onset with low detection rate during the non-movement intention period. Using intracranial EEG (iEEG) from one epileptic subject, we detect movement intention as early as 1500 ms before movement onset with accuracy above 90% using electrodes implanted in the bilateral supplementary motor area (SMA). The coherent results obtained with non-invasive and invasive method and its generalization capabilities across different days of recording, strengthened the theory that self-paced movement intention can be detected before movement initiation for the advancement in robot-assisted neurorehabilitation.

Analysis of Arm Movement Prediction by Using the Electroencephalography Signal

“…Another result which was also shown in the experiment conducted by Lew et al [5] and Planellas et al [8] is the ability of the system to detect the pre-movement intention during "rest" condition. Ideally, for safety reason, the movement intention during "rest" should not be utilized; however, the result of their experiment shows that the system was able to detect movement intentions during "rest" from 10% to 30%.…”

“…This experiment was done using the self-paced method [5][6][7][8]. We illustrate the block diagram of this system which is used in the experiment in Figure 2.…”

“…Experiments conducted by Lew et al [5] and Maoz et al [6] have proven that when using an invasive EEG, the signal will decrease in amplitude before the movement is onset. Moreover, Pirez et al [7], Lew et al [5] and Planellas et al [8] have shown that non-invasive EEG signal can be used to detect pre-movement before the movement is onset, also by detecting the ERD phenomenon. All of the above experiments were conducted using the selfpaced (asynchronous) method, where the subject was given the liberty to move his arm at his own pace within a specific time interval.…”

“…From their experiment, it can be shown that the prediction of movement intention at "rest" condition is below the random level. As it has been proven that by using the ERD phenomenon in time domain [5], classification can be done between the movement intention condition interval or "rest" condition interval; hence, this experiment aims to evaluate the arm movement prediction by using the same phenomenon in the frequency domain with the Burg's autoregressive (AR) method and Support Vector Machine (SVM) before actual movement onset. This experiment also evaluates the implementation of system pattern recognition in continuous time, from 3 seconds before the movement onset until 1 second after the movement onset by using a non-invasive EEG, in order to see the dynamics of system's decision (detection rate) in predicting the arm movement.…”

“…Lew et al [5] tested the system in detecting continuous data 1 second before an auditory cue appears until 2 seconds after the cue appears. From their experiment, it can be shown that the prediction of movement intention at "rest" condition is below the random level.…”