Classification of Mental Effort vs. Relaxation States Using EEG Signals

Electroencephalographic (EEG) signals accurately describe the neural activity of a subject. This paper proposes a classification of EEG signals acquired from 10 subjects into two classes: relaxation and mental effort states. Each subject performs a predefined battery of eight scenario tests, switching between tasks (relaxation and mental effort). To classify the raw EEG signals collected, 12 representative statistical features were extracted and used as inputs for a SVM classifier. To assess the performance of the proposed algorithm, four metrics were computed: Accuracy, Precision, Recall and F1-score. The results confirmed the experimental expectations from a medical perspective and proved that the data are consistent, related to the series of tests, showing that the proposed method can serve as a useful binary classifier, which can discriminate between the relaxation and mental effort states.

I. Introduction

Electroencephalography represents a low-cost, accurate time-resolution method for non-invasively measuring the electrical fields of the brain, using electrodes placed on the scalp. Certain sequences of sinusoidal or sharp, synchronous, or asynchronous, rhythmic, or apparently random waves graphically represent the electrical activity of the brain [1]. The waves' frequency is important because, within signals, it fluctuates according to the condition of the patient and might provide pathological or physiological significance [2]. The Delta wave gives information for assessing the depth of the sleep state, while the Theta wave is specific to certain stages of paradoxical sleep, particularly the Rapid Eyes Moving stage. The Alpha wave manifests in an awake individual exposed to no external stimuli and is typical for the subject at rest. The Beta wave is specific for intense brain activity using several types of sensory stimuli [3].