Advanced Deep Learning for Improving Smart Healthcare Services on Edge and Cloud Computing

Abstract

Cloud computing's capacity to improve the efficiency of smart healthcare services (SHCS) has given it a prominent place in this field. The process of moving reports or data requires too much time and energy, which results in excessive latency and energy problems. In order to ensure preventative treatment and early intervention for those who are at risk, accurate and timely disease prediction is essential. Edge computing offers methods to deal with these problems. The purpose of this study is to suggest an advanced, cloud-enabled, privacy-preserving paradigm that leverages cutting-edge deep learning to enhance healthcare delivery in healthcare organizations. The suggested model successfully recognizes medical entities and is based on a 1-Dimensional-Convolutional Neural Network with Bidirectional Gated Recurrent Unit (1D-CNN-BiGRU). The experimental outcomes reveal that the suggested method performs better state-of-the-art models with F1-score of 99.68%, Recall of 98.96%, a specificity of 98.99%, a precision of 99.46%, and an accuracy of 99.24%, which are significantly better than the current smart healthcare disease prediction systems. Furthermore, by increasing the prediction and diagnosis of health status in clinical practice, the function of edge devices in this SHCS is projected to support clinicians with rapid health-prediction reports via edge servers.