An AI based approach for Medical Decision Support And Critical Alerts

Abstract

Maintaining a long term healthy life condition of chronic patients is an important aspect of medical care. Many human factors contribute to deficiencies and errors in patient care in the local clinical environment. Such as long working hours, high level of stress, lack of enough people may cause to human errors and affecting the effectiveness of the decisions of the physician. In order to reduce this kind of human error rates and to increase the efficiency and accuracy of medical decision making task, This dissertation presents a computer based model we developed as an integrated solution which identifies adverse trend or potentiality of adverse conditions in multiple patient parameter variations, and then provides predictive treatment suggestions based on the likely adverse trends which result in the parameter variations. A combination of neural networks is used for anomaly detection and an expert system is used to provide treatment suggestions. Also final solution has presented as integration of both components. Our model is not an ordinary medical decision support system; it is capable of early identification of adverse trends in medical conditions and providing suitable treatment suggestions, thus acting as a decision support system to assist the treating physician.