AI-Based Medical Diagnosis System for Disease Prediction Using Symtomps

Abstract

Artificial Intelligence (AI) and Machine Learning (ML) have become a prominent part of changing the healthcare sector in the modern world. Machine learning algorithms assist in early detection of diseases by providing an earlier diagnosis, which ultimately leads to a decline in the mortality level, improvement in patient care, and reduction in the cost of treatment. This project introduces an AI-driven medical diagnosis system that predicts illnesses based on the symptoms submitted by the user with the use of machine learning. Symptom-based Analyzed and preprocessed dataset (700+ entries and 17 symptom attributes) was used to train 5 machine learning models: Random Forest, Decision Tree, K-Nearest Neighbors (KNN), Gaussian Naive Bayes, and Logistic Regression. The data was cleaned, coded and visualized to have a view of the distribution of the features of the data and imbalances of classes. The performance of the models was measured on the basis of the metrics accuracy, precision, recall, F1-score, and the confusion matrix. The best prediction accuracy of 93.57 percent in all the models was obtained using KNearest Neighbor (KNN) and was deployed. The last model was embedded into a userfriendly web interface with Streamlit, through which users could enter their symptoms and get an instant diagnosis referring to the disease. The system shows that AI can be useful in making initial medical diagnosis in areas where medical attention is scarce.