Published Paper Details:

Kidney disease poses a significant healthcare challenge worldwide, with millions of patients at risk of chronic complications and kidney failure. Early and accurate detection is essential for improving survival rates and reducing treatment costs. However, conventional diagnostic techniques often rely heavily on manual interpretation, which can lead to variability and delays in diagnosis. To address these limitations, this project introduces a deep learning-based system for automated kidney disease classification. The work integrates advanced workflow management tools--Data Version Control (DVC) for dataset tracking and reproducibility, and MLflow for systematic experiment management, hyperparameter tuning, and model deployment. This integration ensures that the proposed framework is not only accurate but also scalable and transparent, making it suitable for real-world healthcare applications. The core methodology employs a convolutional neural network (CNN) trained on medical imaging data to distinguish between healthy and diseased kidney conditions. The pipeline includes automated data preprocessing, feature extraction, training, validation, and deployment, all while maintaining consistent version control. Experimental results indicate that the system achieves high classification accuracy with reliable performance across multiple test scenarios. Beyond kidney disease detection, this research highlights the potential of combining deep learning models with workflow management tools to create reproducible, maintainable, and adaptable solutions for a wide range of medical imaging challenges.