IIT Madras, CMC Vellore Researchers Develop AI Tools for Early Kidney Disease Detection

The technologies include a machine learning model that predicts chronic kidney disease (CKD) risk using clinical and laboratory data, a deep learning system that analyzes CT scans to classify kidney conditions, and a 3D imaging platform that measures kidney tumor volume and the extent of kidney involvement.

Researchers from Indian Institute of Technology Madras (IIT Madras) and Christian Medical College Vellore (CMC Vellore) have developed three artificial intelligence (AI)-based technologies to support the early detection and assessment of kidney diseases.

The technologies include a machine learning model that predicts chronic kidney disease (CKD) risk using clinical and laboratory data, a deep learning system that analyzes CT scans to classify kidney conditions, and a 3D imaging platform that measures kidney tumor volume and the extent of kidney involvement.

The CT image classifier has been trained on more than 12,000 images and can categorize scans into four groups: normal kidney, kidney cyst, kidney stone, and kidney tumor.

The third technology uses CT scans and open-source software to create 3D models of kidneys. It enables measurement of tumor volume and the percentage of the kidney affected, providing patient-specific information that could support treatment planning.

GL Samuel, professor in the Department of Mechanical Engineering at IIT Madras, said kidney diseases can remain asymptomatic during their early stages and may be detected only after substantial damage has occurred.

“These tools can enable earlier diagnosis, which could help slow down the disease process and reduce the need for expensive interventions like dialysis,” Samuel said.

The CKD prediction model has also been developed as a user-friendly prototype interface to support potential clinical use. Researchers said they are working to improve its accuracy and interpretability for physicians.

Jennifer Delighta, a research scholar at IIT Madras, said early detection could help identify patients at risk and support treatment planning. She added that the patient-specific imaging framework provides a broader assessment of disease extent than standard measurements.

The researchers said the technologies form part of ongoing work toward a kidney Digital Twin, combining AI-assisted imaging with patient-specific 3D anatomical models. The team plans to validate the models using additional patient datasets and explore partnerships with healthcare institutions for real-world deployment.

Future work will also examine integration with minimally invasive wearable sensing systems and Digital Twin platforms for personalized kidney health monitoring.

Stay tuned for more such updates on Digital Health News