AI powered hospital management system (AI-HMS)

Abstract

The increasing complexity of modern healthcare services has highlighted the limitations of traditional Hospital Management Systems, which are primarily focused on the automation of hospital administration and electronic management of patient records, with limited support for decision-making. This thesis proposes the design and development of AI-HMS, an Artificial Intelligence-based Hospital Management System that incorporates the integration of administration and decision making through the use of artificial intelligence. The proposed system is built using a full-stack development framework, with the React.js library for the front-end, the Flask framework for the back-end, and PostgreSQL for the management of both structured and semi-structured data. Supervised learning algorithms are used for the prediction of the risk level of patients and the likelihood of hospital readmission. A Large Language Model-based artificial intelligence assistant is incorporated for the assistance of healthcare professionals and patients. Therefore, AI-HMS is based on the human-in-the-loop philosophy of artificial intelligence, where the artificial intelligence component is used as a decision-support tool. The proposed system was evaluated using synthetic data sets for healthcare services and was found to perform reliably. The system was also deployed as a web-based application, validating the proposed system. Overall, this research demonstrates the potential of using artificial intelligence for the management of hospital services.