NE SCIENCE & TECH BUREAU
CHENNAI, SEPT 3
The kidney often gives no warning until disease has already taken a serious toll. Now, researchers from IIT Madras and CMC Vellore are attempting to give clinicians something medicine has long needed—a faster, data-driven early-warning system that can detect risk, read medical images and map disease with patient-specific precision.
Using AI for earlier kidney disease detection, Researchers from IIT Madras and CMC Vellore have developed three AI-based technologies to support faster and more informed kidney disease assessment.
The team has developed a machine learning model to predict Chronic Kidney Disease… pic.twitter.com/8XP4Nfjqzk
— IIT Madras (@iitmadras) September 3, 2026
Researchers from the Indian Institute of Technology Madras (IIT Madras) and Christian Medical College (CMC), Vellore, have developed a suite of Artificial Intelligence-based tools aimed at supporting the early detection, classification and precise assessment of kidney diseases, potentially enabling physicians to intervene earlier and make more informed treatment decisions.
- Three AI technologies could help doctors detect kidney disease before silent damage becomes irreversible
- Machine-learning model assesses chronic kidney disease risk using clinical and laboratory data
- Deep-learning system trained on more than 12,000 CT images identifies normal kidneys, cysts, stones and tumours
- Open-source 3D platform reconstructs kidneys to measure tumour volume and disease involvement with greater precision
- AI-powered diagnostic support aims to deliver faster, standardised insights and enable earlier clinical intervention
- Research lays groundwork for a kidney “Digital Twin” and future integration with wearable sensors for personalised long-term monitoring
The collaborative research has produced three complementary AI-enabled technologies, bringing together machine learning, deep learning and 3D anatomical imaging.
Together, the tools are designed to address different stages of the diagnostic challenge—from identifying patients at risk of chronic kidney disease to analysing CT scans and precisely assessing the extent of kidney tumours.
Three AI tools, one clinical objective
The first technology is a machine-learning model that uses clinical and laboratory information to predict an individual’s risk of developing chronic kidney disease (CKD).
The second is a deep-learning system trained on more than 12,000 images, capable of automatically analysing CT scans and classifying them into four categories: normal kidney, kidney cyst, kidney stone and kidney tumour.
The third is an open-source 3D anatomical imaging platform that reconstructs kidneys from CT scans and enables precise assessment of tumour volume and the percentage of kidney involvement.
By combining these capabilities, the researchers aim to provide clinicians with rapid and standardised diagnostic support, particularly in fast-paced healthcare environments where timely interpretation of clinical and imaging information can influence treatment planning.
The silent disease problem
Kidney diseases can remain asymptomatic during their early stages, allowing damage to progress before patients are diagnosed.
The researchers believe that AI-assisted risk prediction and imaging analysis could help identify at-risk patients earlier and provide doctors with more consistent information for clinical decision-making.
Earlier detection could, in appropriate cases, enable interventions aimed at slowing disease progression and potentially reducing the need for costly advanced interventions such as dialysis.
However, the researchers position the technology as a clinical decision-support tool, rather than a substitute for medical expertise.
“Quicker and more informed decisions”
The research was led by Prof. G.L. Samuel, Department of Mechanical Engineering, IIT Madras, and Jennifer Delighta, Research Scholar, IIT Madras, in collaboration with Prof. Santosh Varughese, Department of Nephrology, CMC Vellore.
Explaining the research, Prof. G.L. Samuel said:
“The team aimed to develop intelligent systems that would help clinicians make quicker and more informed decisions. We used machine learning along with clinical knowledge to develop tools that would assist in the earlier detection of kidney diseases and give more detailed information specific to the patient.”
The CT image-classification system has been trained using more than 12,000 images and is designed to distinguish healthy kidneys from cysts, stones and tumours.
Meanwhile, the 3D imaging framework, developed using open-source software, offers a potentially inexpensive and repeatable method of measuring tumour burden—information that could assist clinicians in assessing disease extent and planning treatment.
From a scan to a patient-specific 3D kidney
Jennifer Delighta highlighted the importance of early diagnosis and the patient-specific nature of the imaging technology.
“Early detection is of paramount importance when dealing with kidney diseases; these AI tools can help detect at-risk patients early and plan their treatment more effectively. The patient-specific imaging framework is of significant promise as it goes beyond the standard measurements to give a more comprehensive picture of the extent of the disease.”
The CKD prediction model has also been implemented as a user-friendly prototype interface, with the researchers working to improve its accuracy and interpretability for doctors who may use its predictions in clinical settings.
Towards a ‘digital twin’ of the kidney
Perhaps the most futuristic dimension of the research is its potential contribution to a kidney Digital Twin.
The concept involves combining AI-assisted image analysis with patient-specific 3D anatomical models, creating a virtual representation that could eventually support personalised clinical decision-making.
Such Digital Twin platforms could, subject to further validation, potentially help clinicians monitor disease, forecast changes and plan individualised therapeutic strategies.
The research received institutional support from IIT Madras and the SPARC (Scheme for Promotion of Academic and Research Collaboration) project.
From lab validation to real-world healthcare
The researchers now plan to test the models using larger and more diverse patient datasets and undertake multi-centre clinical validation.
The team also intends to strengthen partnerships with healthcare institutions to assess how the technologies can be translated into real-world clinical environments.
Looking further ahead, researchers are exploring integration of the AI technologies with minimally invasive wearable sensing systems and Digital Twin platforms for long-term, personalised kidney-health monitoring.
The larger vision is compelling: a future in which a patient’s clinical data, medical images, anatomical model and continuous health signals could work together to give doctors a more comprehensive picture of kidney health.
For a disease that can remain silent while damage accumulates, the ability to detect risk earlier, see disease more clearly and understand each patient more precisely could prove as important as the treatment itself.
