Internet-of-Things-Assisted Wireless Body Area Network-Enabled Biosensor Framework for Detecting Ventilator and Hospital-Acquired Pneumonia

Ventilator-associated pneumonia (VAP) and hospital-acquired pneumonia (HAP) are the leading cause of death in intensive care units (ICUs) developed two days after endotracheal intubation and hospitalization or ICU admission. Hospital-acquired pneumonia affects ventilated patients twice as frequently as nonvented patients. Detecting volatile organic compounds (VOCs) in exhaled breath can differentiate between sick and healthy people. A noninvasive biosensor framework is necessary to detect VOC-induced pneumonia from reducing mortality rates in ICUs. To identify symptoms of pneumonia, researchers have developed a portable and wearable biosensor arrays and machine learning frameworks to examine VOCs in exhaled air. Wireless body area networks (WBANs) can allow ubiquitous devices and internet-enabled ICU monitoring for health tracking. These findings suggest that a machine learning system built by biosensors and the Internet of Things (IoT) can recognize pneumonia contracted in hospitals and ventilators. A 128-core NVIDIA Jetson Nano graphics processing unit (GPU) enables the seamless transmission of VOC data and other patient biological characteristics to the Amazon Web Service (AWS) IoT Core. The developed support vector machine (SVM) and k-nearest neighbor (KNN) framework are deployed in Jetson Nano, and the SVM model outperforms the KNNs model in terms of accuracy (92.35%), sensitivity (92.67%), precision (93.38%), and receiver operating characteristic (ROC, 93.11%).