User-adapted neurostimulation for stress reduction

Acute stress is a physiological and psychological condition that manifests when the demands of either a real-life situation or a cognitive task exceed the resources we have available to cope. If not treated, acute stress can turn into chronic, resulting in serious physical and psychological issues, including cardiovascular diseases, anxiety, and decreased productivity. Furthermore, it may have significant economic consequences, such as increased absenteeism in the workplace. This project aims to develop a real-time algorithm for stress detection and reduction by exploiting binaural beats. Binaural beats are an auditory stimulation that occurs when two tones with a slight frequency mismatch are presented separately to each ear. This results in a perceived signal, reconstructed in our brain, with an amplitude that oscillates at a frequency equal to the difference between the two tones. While existing literature is mainly focused on fixed stimulation approaches with contradictory results, this thesis aims to evaluate whether an adaptive stimulation approach, piloted by a stress detection algorithm, can effectively reduce stress. In other words, the goal is to identify a personalised stimulation frequency in the theta band (4 to 8 Hz) to induce relaxation. An experimental protocol to induce acute mental stress was designed, considering as stressor a mental arithmetic task with a limited time to provide the answers. In addition, different questionnaires were filled out by the volunteers to analyse the psychological responses and the affective conditions of the participants (18 undergraduate students). The technique was compared with fixed and no-stimulation conditions. For this study, the PolarH10 chest strap was used as a device for data collection. The features from both the electrocardiogram (ECG) and the breath rate, indirectly estimated from the ECG, were considered for our method. By doing so, the connectivity between respiratory activity and the heart is considered. Results show that the experimental protocol was capable of inducing stress in the participants. Indeed, the heart rate is higher during the task than during the baseline (p<0.001), i.e., a period of 2 minutes before the beginning of the cognitive task, which lasts 4 minutes. Furthermore, we detected a significantly lower value of the regressor’s output (p=0.022) when comparing the adapted stimulation approach with the fixed stimulation condition. No statistical differences were detected in terms of accuracy and reaction times in the cognitive tasks, indicating no learning effect by the participants. Future studies are required to assess the goodness of the proposal, both considering additional candidates and different experimental conditions.