Fusion-driven EEG reconstruction and cognitive workload recognition using conditional diffusion and graph-based learning

Many studies have shown that most maritime accidents are attributed to human error as the initiating cause, resulting in a need for study of human factors to improve safety in maritime transportation. Among the various techniques, Electroencephalography (EEG) has the key advantage of high time resolution, with the possibility to continuously monitor brain states including human mental workload, emotions, stress levels, etc. In this paper, we proposed a novel mental workload recognition algorithm using deep learning techniques that outperformed the state-of art algorithms and successfully applied it to monitor crew members’ brain states in a maritime simulator. We designed and carried out an experiment to collect the EEG data, which was used to study stress and distribution of mental workload among crew members during collaborative tasks in the ship’s bridge simulator. The experiment consisted of two parts. In part 1, 3 maritime trainees fulfilled the tasks with and without an experienced captain. The results of EEG analyses showed that 2 out of 3 trainees had less workload and stress when the experienced captain was present. In part 2, 4 maritime trainees collaborated with each other in the simulator. Our findings showed that the trainee who acted as the captain had the highest stress and workload levels while the other three trainees experienced low workload and stress due to the shared work and responsibility. These results suggest that EEG is a promising evaluation tool applicable in human factors study for the maritime domain.

Mental workload can be recognized from Electroencephalogram (EEG) and can be used to assess mental efforts of the user performing different tasks. In this work, we designed and implemented an experiment for mental workload recognition related to no-task, visual task, auditory task and multitask performance. The Simultaneous Capacity SIMKAP test was used to induce different levels of mental workload related to multitasking in 12 subjects. EEG data was collected with Emotiv device, processed and analyzed using power, statistical, fractal dimension (FD) features with Support Vector Machine (SVM) and k-Nearest Neighbors (k-NN) classifiers. The best accuracy of 90.39% for 2 classes and 80.09% for 4 classes using SVM was achieved when statistical and FD feature combination was used. The proposed algorithm can be applied for mental workload monitoring.

Towards an effective cross-task mental workload recognition model using electroencephalography based on feature selection and support vector machine regression