AI-Powered Disease Diagnosis from Medical Image Reports
L R Kaviraj, R. Rajalakshmi, S Joshua
Abstract
The use of data science and technology in healthcare has completely changed how diseases are predicted and diagnosed. Through the integration of two state-of-the-art deep learning architectures—InceptionV3, VGG 19 and sequential CNN within a Flask-based API framework, this research offers a novel method for multi-disease prediction. Kidney stones, kidney cysts, kidney tumors, Lung cancer, and brain tumors are among the disorders that are being targeted. The project progresses through a number of crucial phases. first, a wide-ranging and extensive dataset made up of image data that was rigorously preprocessed. This includes data augmentation, variable normalization, and handling missing data. Modern deep learning architectures, such as Transfer learning models like InceptionV3 and EfficientNet-B7, are chosen and optimized for each disease in order to extract complex patterns from the medical imaging data. intricate connections among high-dimensional data, improving the precision of predictions. These structures are well known for improving prediction accuracy by capturing intricate linkages within high-dimensional data. The performance of the model is evaluated by calculating accuracy and loss after training.