Kidney Disease Classification Using Mlflow
Keywords:
Convolutional Neural Networks (CNN), MLflow, kidney disease detection, healthcare efficiency.Abstract
With the increasing prevalence of kidney-related health issues, timely and accurate diagnosis has become a
critical concern in the medical field. This project proposes a deep learning-based classification system for
kidney disease using medical imaging and automated model tracking tools. Leveraging a custom dataset and
MLflow integration, the workflow involves data preprocessing, CNN-based model training, evaluation, and
version-controlled deployment. Feature configurations are managed through modular configuration files,
and metrics such as accuracy, loss, and evaluation scores are monitored and logged in real-time.
Models including Convolutional Neural Networks (CNN) were trained and evaluated, with high accuracy
demonstrating the model's potential for clinical application. The integration of MLflow streamlined
experiment tracking and ensured reproducibility, while the modular design allowed for scalable and
maintainable experimentation. Overall, the research establishes a reliable foundation for future AI-powered
diagnostic tools aimed at improving kidney disease detection and healthcare efficiency.
Downloads
References
1. Alelign & Petros (2018). Kidney stone disease update.
2. Caglayan et al. (2022). DL-assisted kidney stone detection on CT. Int. Braz. J.
3. Sabuncu et al. (2023). Kidney stone classification using CT images. Biomed. Tech.
4. Jyotismita & Ayşegül (2024). Ensemble DL networks for detection.
5. Sassanarakkit et al. (2022). ML in kidney stonemanagement.
6. Leube et al. (2023). PSMA-PET for enhanced kidney segmentation. Z. Med. Phys.
7. Junyu et al. (2023). Style transfer for MRI segmentation. Radiology
8. Gaikar et al. (2022). Transfer learning for MRI kidney segmentation.
9. Felix et al. (2022). Image processing-based stone detection.
10. Angshuman et al. (2022). Holistic approach for kidney stone detection.
11. Li et al. (2022). DL networks for stone detection & segmentation. Diagnostics
12. Amiri et al. (2021). Radiomics & ML for kidney damage prediction.
13. Ma et al. (2020). CKD detection using hybrid ANN. Future Gen. Comput. Syst.
14. Pande & Agarwal (2024). Multi-class kidney abnormality detection.
15. Saif et al. (2024). Deep-kidney: DL for CKD prediction.
16. Subedi et al. (2023). Kidney CT classificationvia Vision Transformer.
