Kidney Disease Prediction using Deep Learning

Abstract

Early and accurate detection of kidney-related diseases is vital for timely intervention and treatment. This project introduces a deep learning–based Kidney Disease Prediction System that identifies four major conditions—Cyst, Stone, Tumor, and Normal—from medical images. The system is built using Convolutional Neural Networks (CNNs) and MobileNet, both trained to classify renal abnormalities with high accuracy. A dynamic and user-friendly GUI, developed with PyQt5, allows users to enter patient details, upload diagnostic images, and receive predictions alongside medical descriptions, symptoms, and treatment recommendations. Furthermore, the application integrates a SQLite database to store patient records and automate report generation. This comprehensive solution enhances diagnostic precision, reduces human error, and offers healthcare professionals an efficient tool for real-time disease screening and decision support.