Anti-Heartbeat-Evoked Potentials Performance in Event-Related Potentials-Based Mental Workload Assessment

Area of Science:

• Neuroscience
• Cognitive Science
• Biomedical Engineering

Background:

• Event-related potentials (ERPs) are crucial for understanding cognitive processes, including mental workload (MWL).
• Heartbeat-evoked potentials (HEPs) are physiological signals that can potentially contaminate neural recordings.

Purpose of the Study:

• To investigate the impact of heartbeat-evoked potentials (HEPs) on the accuracy of classifying mental workload (MWL) using event-related potentials (ERPs).
• To compare MWL classification performance when HEPs are included versus excluded from ERP analysis.

Main Methods:

• Participants performed a mental arithmetic task to induce low and high mental workload (MWL) states.
• Electroencephalography (EEG) was used to record event-related potentials (ERPs).
• Trials were categorized into three conditions: ERP_HEP (heartbeat present), ERP_A-HEP (heartbeat absent), and ERP_T (all trials).
• A radial basis function-support vector machine with 10-fold cross-validation was employed to classify MWL based on P600 amplitude and latency.

Main Results:

• The ERP_A-HEP condition, excluding heartbeats, achieved 100% accuracy in MWL classification.
• This represents a significant improvement compared to the ERP_T condition (85.7% accuracy) and the ERP_HEP condition (71.4% accuracy).
• The presence of HEPs demonstrably reduced the accuracy of MWL classification.

Conclusions:

• Heartbeat-evoked potentials (HEPs) overlap with and interfere with event-related potentials (ERPs), diminishing their utility for cognitive state classification.
• Excluding HEPs from ERP analysis significantly enhances the performance of mental workload (MWL) classification algorithms.
• This study highlights the importance of considering cardiac artifacts in neurophysiological research, particularly for ERP-based cognitive workload assessment.