Automated Patient Face Registration and Wait Time Prediction Using AWS with LSTM

The current healthcare services are inadequate for quick access in terms of patient registration, access recognition, and estimation of wait times. It is more time-consuming and less efficient. To overcome these obstacles, the project is designed in aiming to implement an automated system with cloud services for patient registration, facial recognition, and wait time prediction. This automated service would enable clinics and hospitals to provide quick access for patients in terms of registration, access recognition, and wait time prediction. The system incorporates AWS tools to develop the platform. It includes tools such as Rekognition, Lambda, to register patients, face recognition, and wait time prediction based on historical data. It incorporates Long Short-Term Memory to improve the accuracy of wait time prediction. Captured incoming images of patients are used for processing by AWS Lambda using Rekognition either as a way to register new or recognize patients based on their face IDs. The face IDs registered are displayed by the DynamoDB, whereas the classification of data and wait times will be predicted using pre-fed and real-time data through Amazon SageMaker. The system reduces manual intervention and increases accuracy; therefore, it cuts down the time of patient registration, increasing the healthcare workflow's effectiveness. On the whole, the system is a one-piece solution for manual intervention and provides quick access for patients to healthcare services. Experimental results show the possibility of the system for timely provision of information to the patients and reasonable estimates of wait times.