IoT-assisted irregular environmental event determination for health analysis of pregnant females

Indoor environment pollution is one of the significant concerns related to the individual's health of every category. As a majority of the population is spending their time indoors, the research on environment monitoring is realized on priority. In this research, the correlation between the indoor environment and health is determined to predict the health severity of pregnant females in real-time. The proposed solution employs IoT-enabled smart sensors for data acquisition. The accumulated data is processed over the fog-cloud platform for data quantification. To classify the irregular events from the processed data, the machine learning-assisted Weighted-Naive Bayes (W-NB) classification technique is deployed on the fog layer. Furthermore, a deep learning-inspired multilayered convolutional neural network-long short-term memory (M-CNN-LSTM) model is proposed to determine the scale of health severity. Moreover, a two-phased alert generation mechanism is proposed to deal with the case of a medical emergency by generating notifications to the concerned medical specialists. The performance of the system is validated by performing several experimental simulations over the environmental and health datasets comprised of nearly 38 356 instances. The performance of the proposed solution is evaluated in terms of event classification, severity prediction, temporal delay efficiency, and overall performance stability. The calculated outcomes registered considerable performance over the several state-of-the-art decision-making approaches.

Automated summary

This study focuses on monitoring indoor environmental pollution using smart sensors and IoT technology to predict the health status of pregnant women and proposes a solution to address health emergencies. Experimental simulations on environmental and health datasets validate the system's performance in event classification, severity prediction, temporal delay efficiency, and overall performance stability. The proposed solution demonstrates considerable performance compared to existing decision-making approaches.