A new framework to enhance healthcare monitoring using patient-centric predictive analysis

Abstract

In the contemporary healthcare landscape, various intelligent automated approaches are revolutionizing healthcare tasks. Learning concepts are pivotal for activities like comprehending acquired data and monitoring patient behavior. Among patient-centric concerns, addressing data heterogeneity, extraction, and prediction challenges is crucial. To enhance patient monitoring using care indicators like cost and length of stay at healthcare centers, many researchers found a model for automated tools, but do not have the artificial intelligence (AI) based models as of now. Therefore, this research study will propose an AI and internet of things (IoT) integrated automated approach with smart sensors called the “PatientE” framework with heterogeneity and patient data. Employing certain rules for data extraction to form a distinct representation, our model integrates pre-treatment information and employs a modified combined random forest, long-short term memory (LSTM), and bidirectional long-short term memory (BiLSTM) algorithm for predictive post-treatment monitoring. This framework, synergizing AI, IoT, and advanced neural networks, facilitates real-time health monitoring, especially focusing on breast cancer patients. Embracing pre-treatment, in-treatment, and post-treatment phases, our model aims for accurate diagnosis, improved cost-efficiency, and extended stays. The evaluation underscores scalability, reliability enhancement, and validates the framework's efficacy in transforming healthcare practices.