Applied Difference Techniques of Machine Learning Algorithm and Web-Based Management System for Sickle Cell Disease

摘要

Machine learning algorithm and web-based application systems have played a major role in improving the healthcare organisation in terms of continuous tele-monitoring therapy and maintaining telemedicine management systems. Currently, no intelligent system has been used in terms of managing sickle cell disease. However, this paper presents a system that facilitates a shift from manual methods to automated approach that can offer fast data collection with a reduced error rate. The system will be able to examine patient data and provide a suitable amount of Hydroxycarbamide drugs/liquid for each patient. By using a web-based system, we tend to improve patient welfare and mitigate patient illness before it gets worse over time, particularly with elderly people. The system will forward any critical concerns from the patient, it generates an automatic message to the medical doctors based on web-based platform in order to assist them with optimal decision-making. The initial case study that has been addressed in this project is how to make predictions for sickle cell disease for the amount of dose based on different architectures of machine learning in order to obtain high accuracy and performance. The most significant key for making predictions of the amount of medication is to enable healthcare organisation to provide accurate therapy recommendations based on previous data. The results using ANN showed that the proposed network produces significant improvements using the different evaluation methods. In our experiments, The MLP algorithm produced the best result in terms of the lowest error rates compare with other techniques. The Mean Square Error, Root Mean Square Error, Mean Absolute Error, and Mean Absolute Percentage Error achieved 17887.55, 133.74, 90.20 and 0.1345, respectively.