.png){kind=icon}
.png){kind=icon}
Return
Adaptive EEG channel phase synchronization for individualized mental fatigue detection
.png){kind=icon}
PRE
AI
J
C
X
DOI：10.1016/j.bspc.2026.111218.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
Mental fatigue is associated with reduced health and work efficiency, making reliable detection important. However, existing EEG-based methods are often challenged by noise contamination, inter-subject variability, and limited adaptability. This study proposes a modified enhanced singular value decomposition (mESVD) framework together with an adaptive EEG channel phase synchronization method to address these issues. The mESVD framework dynamically determines the reconstruction order based on singular envelope characteristics, enabling noise reduction while preserving task-relevant EEG patterns. Experimental results indicate that the proposed framework provides more stable feature representations and yields improved classification performance compared with conventional preprocessing approaches. The adaptive phase synchronization method identifies subject-specific relevant EEG channels using weighted phase lag index (wPLI) scores and a ranking strategy. This individualized channel selection approach reduces the influence of inter-subject variability and results in an average classification improvement of 7.05 % relative to comparison models. Connectivity analysis suggests a fatigue-related redistribution of functional connectivity patterns. Compared with the non-fatigue condition, fatigue is associated with reduced long-range connectivity and relatively increased right-hemisphere involvement, particularly in frontal and frontotemporal regions. Overall, the proposed framework provides a systematic approach for improving EEG preprocessing and adaptive channel selection in mental fatigue recognition.
Journal
IF:
4.9
Papers:
9.7K
Citations:
2.4W
Organization
No organization information available
Cited Papers