Real-time Neonatal Monitoring Using a Hybrid Deep Learning Model

The neonatal intensive care unit (NICU) provides emergency medical care for preterm babies. In some of the existing NICUs, neonatal parameters are monitored and analyzed by healthcare professionals. This can lead to human errors and further delays during emergency cases. The emergence of Internet of Things (IoT) devices has improved real-time data collection in NICUs. IoT devices generate enormous data every second in the NICUs. Therefore, it is important to design an efficient artificial intelligence algorithm to analyze the data in the NICUs. Among artificial intelligence, deep learning model excels in identifying the features of data. A hybrid deep learning model integrates two or more distinct deep learning models to enhance the overall performance. This article introduces a hybrid deep learning model that integrates an attention mechanism, a long short-term memory (LSTM) network, and a one-dimensional convolutional neural network (1D-CNN) for classifying NICU data. A simulation of the proposed model has been conducted using Python. The proposed model obtained a validation accuracy of 97.44%, a validation loss of 0.0794, and an F1-score of 0.98. The simulation results demonstrate that the model efficiently classifies the NICU data.