Sign in Subscribe
Machine learning

AI-Powered Diagnosis for Chronic Kidney Disease Breakthrough: Daffodil International University's Research on Personalized Medicine

Strategic Insights

— 1 min read

Researchers from Daffodil International University have developed a novel AI-powered system for early prediction and classification of chronic kidney disease (CKD), revolutionizing personalized medicine for patients worldwide. The team leveraged machine learning algorithms and ensemble methods to analyze clinical features of CKD, achieving unprecedented accuracy in diagnosis. The breakthrough has significant implications for the healthcare industry, as the integrated system can facilitate real-time monitoring, predictive analytics, and efficient CKD diagnosis.

Key Takeaways:

  • The research, published in Discover Applied Sciences, utilized six base machine learning classifiers and two novel ensemble models (MKR Stacking and MKR Voting) to predict and classify CKD.
  • MKR Stacking achieved the highest accuracy of 99.50%, outperforming Random Forest (98.75%) and MKR Voting (98%).
  • The proposed AI-powered system integrates high-performing models into the Internet of Medical Things and Robotic Process Automation frameworks for real-time monitoring and predictive analytics.
  • The research advocates for integrating machine learning applications into CKD management, enabling early interventions and personalized treatment plans.
  • Md. Hasan Imam Bijoy, Md. Jueal Mia, Md. Mahbubur Rahman, Mohammad Shamsul Arefin, Pranab Kumar Dhar, and Tetsuya Shimamura were among the researchers involved in the study.

Statistics:

  • Accuracy of MKR Stacking: 99.50%
  • Accuracy of Random Forest: 98.75%
  • Accuracy of MKR Voting: 98%
  • Computational time for MKR Stacking: [Not specified]
  • Number of CKD datasets used in the study: 5
  • Number of robust feature selection techniques used: 4 (Lasso, Fisher score, Information Gain, and Relief)
  • CKD website: [Not specified]
  • Project title: [Not specified]

Sources:

  • A robot process automation based mobile application for early prediction of chronic kidney disease using machine learning. Discover Applied Sciences, 2025,7(6):1-34. Publisher: Springer.
  • NewsRx. Researchers at Daffodil International University Zero in on Personalized Medicine (A robot process automation based mobile application for early prediction of chronic kidney disease using machine learning). Health & Medicine Week. June 13, 2025; p 5630

Read more