A multi-sensor based emergency healthcare monitoring system integrating heart status, stress, and alcohol detections

Abstract

PurposeUbiquitous health-care monitoring systems can provide continuous surveillance to a person using various sensors, including wearables and implantable and fabric-woven sensors. By assessing the state of many physiological characteristics of the patient’s body, continuous monitoring can assist in preparing for the impending emergency. To address this issue, this study aims to propose a health-care system that integrates the treatment of the impending heart, stress and alcohol emergencies. For this purpose, this study uses readings from sensors used for electrocardiography, heart rate, respiration rate, blood alcohol content percentage and blood pressure of a patient’s body.Design/methodology/approachFor heart status, stress level and alcohol detection, the parametric values obtained from these sensors are preprocessed and further divided into four, five and six phases, respectively. A final integrated emergency stage is derived from the stages that were interpreted to examine at a person’s state of emergency. A thorough analysis of the proposed model is carried out using four classification techniques, including decision trees, support vector machines, k nearest neighbors and ensemble classifiers. For all of the aforementioned detections, four metrics are used to evaluate performance: classification accuracy, precision, recall and fmeasure.FindingsEventually, results are validated against the existing health-care systems. The empirical results received reveal that the proposed model outperforms the existing health-care models in the context of metrics above for different detections taken into consideration.Originality/valueThis study proposes a health-care system capable of performing data processing using wearable sensors. It is of great importance for real-time systems. This study assures the originality of the proposed system.