MediSmart: An Integrated AI Framework for Hospital Management and Patient Safety

Department of Computer Engineering, Vishwakarma Institute of Technology, Bibwewadi, Pune, India

^* Corresponding author: This email address is being protected from spambots. You need JavaScript enabled to view it.

Abstract

Hospitals manage an array of clinical activities, from assessing patient symptoms accu- rately to prescribing safe medications and managing clinical resources efficiently. Manually managing these activities in a hospital can cause delays in providing treatments, making medication errors, and not utilizing beds correctly in the hospital. MediSmart was designed using a unified AI-based approach that combines medication recommendations, drug-drug interactions detection and bed occupancy fore- casting. Medication recommendations will be made using a Random Forest classification process, and harmful drug interactions will be identified using a hybrid of rule-based and machine learning algo- rithms. Bed occupancy will be forecasted using Long Short-Term Memory (LSTM) Networks. The experiments conducted on the simulated hospital data show that the accuracy of the medication recommendations was 92%, the accuracy of detecting adverse drug events was 88%, and the accuracy of forecasting bed occupancy was 89%. As such, the MediSmart concept has the potential to create a safer clinical environment, reduce the incidence of human error, and promote the use of data to support proactive management of hospital operations.