Development of a Polymeric Optical Fiber Sensor for Stress Estimation: A Comparative Analysis Between Physiological Sensors

The global prevalence of stress and its far-reaching impact on well-being has spurred the pursuit of innovative solutions for physiological and stress monitoring. Existing methods, whereas informative, often lack cost-efficiency, daily usefulness, real-time capabilities, or user comfort. This study introduces a novel approach using a low-cost wearable polymeric optical fiber (POF) sensor to classify stress in 12 healthy individuals (seven women and five men) using the Trial Social Stress Test (TSST) method for stress induction. The study’s methodology involves validating the POF sensor’s physiological measurement, comparing it with a fiber Bragg grating (FBG)-based sensor and with an electrocardiogram (ECG) commercial sensor, and subsequent making a stress classification approach using a Bagged Decision Tree Classifier (BDTC). To the best of the authors’ knowledge, this is the first work implementing POF and FBG sensors in stress detection. The results showcase the POF sensor’s proficiency in capturing pulse and respiration signals, aligning closely with established monitoring sensors. The tested classification algorithm with the POF sensor exhibits an accuracy of 92.37% in the validation stage and 84.75%, recall of 79.41%, precision of 93.10%, and F-score of 85.71%. These results demonstrate the POF sensor’s potential to be a reliable stress indicator. In addition, considering the reduction in the size of the optical interrogation system for the POF sensor compared with FBG interrogation systems, the proposed sensor can be used in complete wearable systems. For future work, the sensor will be improved in terms of technology readiness level (TRL) and tested in real-scenario applications.