<div><p>In this work, an EEG-based control
paradigm assisted by micro-facial-expressions (microFE-BCI) was developed, focusing
on the mainstream defect as the insufficiency of real-time capability,
asynchronous logic, and robustness. The core algorithm in microFE-BCI contained
two stages (asynchronous ‘ON’
detection & microFE-BCI
based real-time control) with four steps
(obvious non-microFE-EEGs exclusion,
interface ‘ON’ detection,
microFE-EEGs real-time decoding,
and validity judgment).
It provided the asynchrounous function, decoded 8 instructions from the latest
100 ms EEGs, and greatly reduced the frequent misoperation. In the offline
assessment, microFE-BCI achieved 96.46%±1.07 accuracy for interface
‘ON' detection and 92.68%±1.21
for microFE-EEGs real-time decoding,
with the theoretical output timespan less than 200 ms. This microFE-BCI was
implemented into a software, and applied to two online manipulations for
evaluating the stability and agility. In object-moving
with a robotic arm, the averaged IoU
was 60.03±11.53%. In water-pouring with
a prosthetic Hand, the averaged
water volume was 202.5±7.0 ml. During online, microFE-BCI performed no
significant difference (P = 0.6521 & P = 0.7931) with commercial control
methods (i.e., FlexPendant and Joystick), indicating a similar level of
controllability and agility. This study demonstrated the capability of
microFE-BCI, enabling a novel solution to the noninvasive BCIs in real-world
challenges.</p></div>