Virtual Medical Assistant System for Diseases Detection using Machine Learning

Health is very important for human lives. The health sector is growing day by day. Nowadays, due to information and communication technology, the health sector has grown immensely. Therefore, the virtual medical assistant system is developed by combining the machine learning algorithm to become part of this growth. The Virtual Medical Assistant System predicts heart disease according to the health parameters like age, height, weight, BMI, and hypertension given by the user. For pattern recognition and classification problems, K-Nearest Neighbor (k-NN), back-propagation neural networks, Naive Bayes, and Support Vector Machines (SVM) algorithms are employed to implement the system. The SVM classifier performs the best for the classification problems, and XGBoost is combined for feature extraction purposes to develop new feature combinations for training the SVM model. And from the experimental analysis, the accuracy obtained for SVM is 90.S%, but for others it’s less. This proposed system checks, whether a person is affected by the disease or not and recommends the right specialist for that disease, which is a difficult process nowadays. So it saves the patient both time and money. This work is to provide a virtual medical assistant that can analyze symptoms, predict diseases, and recommend a specialized doctor.