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 (ICU) developed two days after endotracheal intubation and hospitalization or intensive care unit admission. Hospital-acquired pneumonia affects ventilated patients twice as frequently as non-vented patients. Detecting volatile organic compounds (VOCs) in exhaled breath can differentiate between sick and healthy people. A non-invasive biosensor framework is necessary to detect VOC-induced pneumonia from reducing mortality rates in intensive care units. 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 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 AWS IoT Core. The developed SVM and KNN framework is deployed in Jetson Nano, and the SVM model outperforms the K Nearest Neighbors model in terms of accuracy (92.35%), sensitivity (92.67%), precision (93.38%), and receiver operating characteristic (ROC,93.11%).