How a UB graduate student is using AI to transform kidney transplantation

When surgeons at Peter Bent Brigham Hospital in Boston performed the first-ever kidney transplant in 1954, they were unaware that the procedure would revolutionize health care and make it possible to treat diseases once considered incurable.

Organ transplantation is a complex surgical procedure that involves replacing a failing organ with a healthy one from someone else who doesn’t need it.

“Being a donor is one of the most selfless acts a person can do. It’s a way to turn loss into hope and give others the gift of life,” said Kymani Getfield, a senior nuclear medicine technology major and UB CSTEP campus health educator.

In many cases, patients undergoing kidney transplantation suffer from comorbidities — the presence of two or more diseases at the same time — which not only causes significant discomfort but also forces them to meet with multiple medical professionals to develop detailed dietary and lifestyle plans suited to their conditions.

When Nithin Rogan Aralsumy, a graduate student in industrial engineering, learned about the challenges and stress patients faced while trying to receive these personalized lifestyle plans, he set out to make the process smoother, more efficient and less burdensome by creating NutriNephraAI.

“Our project focuses on using agentic AI architecture with knowledge graphs to deliver personalized diet plans for patients with chronic kidney disease who also have comorbidities,”  Rogan said.

NutriNephraAI uses Agentic AI, a type of artificial intelligence capable of acting independently without constant human intervention. The implementation of this system allows for a holistic approach to developing lifestyle plans for patients, reducing the need for multiple visits to different medical specialists.

The process of receiving a personalized plan can be completed in five simple steps. First, users upload their clinical reports, which detail their comorbidities. Next, the NutriNephraAI model automatically analyzes the report and generates a tailored set of clinically relevant questions based on priority conditions. The system then produces context-aware, evidence-based responses for each question, providing dietary guidance and mitigation strategies. Finally, correlations, risk percentages and abundance thresholds are calculated to create a comprehensive report focused on dietary recommendations, intervention strategies and risk mitigation.

“The diet recommendations are primarily tailored to improve microbiome flora. Each person has a unique microbiome flora, and patients with conditions such as chronic kidney disease often have irregular flora,”  Rogan said.

When asked about the primary motivation behind creating NutriNephraAI,  Rogan shared that long wait times for simple consultations played a major role.

“Before NutriNephraAI, I worked in health care, and a lot of patients would experience long wait times for simple consultations, especially if they needed to meet with nutritionists for their conditions,” he said.

Chronic Kidney Disease rose from the 16th to the ninth-leading cause of death worldwide, and in many cases, kidney transplantation is the only life-sustaining treatment. In areas where quality medical care is absent, NutriNephraAI can help provide expert advice without the need to visit medical professionals in person.

“A lot of patients don’t have access to quality health care and often struggle to make doctor appointments, which is where our platform comes in,”  Rogan said.

With NutriNephraAI, Rogan is helping bridge the gap between technology and patient care, making personalized health care more accessible for those living with chronic conditions. His platform could lead to innovations in organ transplantation and help patients receive expert recommendations without the wait times.

The features desk can be reached at [email protected]