Mount Sinai Secures $4M NIH Grant to Develop AI Model for Predicting Postoperative Acute Kidney Injury

Mount Sinai Receives $4 Million NIH Grant for Postoperative Acute Kidney Injury Research

Researchers at the Icahn School of Medicine at Mount Sinai have secured a five-year National Institutes of Health grant of approximately $4 million to advance prediction and prevention of postoperative acute kidney injury (AKI), a serious complication that affects an estimated 12 percent of the 50 million major noncardiac surgeries performed annually in the United States1

1

2

2

. This initiative addresses a critical gap in surgical care, as postoperative acute kidney injury can extend hospital stays and, according to previous Mount Sinai-led research, increase the risk of chronic kidney disease and death. With no direct treatment currently available, identifying high-risk patients and implementing preventive measures become essential.

Understanding the Risk Factors for Acute Kidney Injury After Surgery

During and after surgery, patients experience significant physiological stress, including fluctuations in blood pressure, heart rate, breathing patterns, and potential bleeding. The kidneys prove particularly sensitive to these changes, making them among the first organs to show signs of injury1

1

. The risk escalates for patients with pre-existing kidney or heart conditions, those undergoing major surgical procedures, or individuals experiencing other complications. "Postoperative acute kidney injury is a common and serious complication after surgery, yet we still lack reliable ways to identify which patients are most likely to develop it and determine how best to prevent it," says Andrew Leibowitz, MD, Chair of the Department of Anesthesiology, Perioperative and Pain Medicine at the Icahn School of Medicine at Mount Sinai2

2

.

Developing an AI Model Predicting Postoperative AKI Using Federated Learning

The Mount Sinai-led team will analyze data from more than 320,000 patients at 15 hospitals across three health systems to develop and test an AI model predicting postoperative AKI1

1

. The research brings together experts from Mount Sinai, Vanderbilt University, and UCLA, combining expertise in anesthesiology, nephrology, biomedical informatics, and health data science. The team will employ federated learning, an innovative method that allows hospitals to develop a shared predictive tool for AKI risk without transferring individual patient records between institutions. Each hospital retains its own decentralized patient data, addressing privacy concerns while enabling collaborative model development2

2

.

Leveraging Electronic Health Records for Personalized Prevention Strategies

"Today, we have limited ability to predict who will develop kidney injury after surgery, and prevention strategies are often applied broadly rather than tailored to individuals," explains Ira Hofer, MD, Associate Professor of Anesthesiology, Perioperative and Pain Medicine at Mount Sinai and lead investigator for the study. "Our goal is to use the wealth of information available in electronic health records to identify patients at risk before surgery and move toward prevention strategies that are better matched to each patient"2

2

. This AI-driven analysis represents a shift from one-size-fits-all approaches to personalized prevention strategies based on individual patient characteristics and risk profiles.

Addressing Limitations of Existing Prediction Models

Many existing postoperative acute kidney injury prediction models were developed with relatively small patient populations at individual hospitals, making it difficult to determine whether a model will work reliably at other hospitals or for different patient populations1

1

. "Federated learning gives us a way to learn from a much larger and more diverse patient population while keeping patient-level data within the health systems where it originated," says Girish N. Nadkarni, MD, MPH, Chair of the Windreich Department of Artificial Intelligence and Human Health at Mount Sinai. "That allows us to ask whether a prediction model can work reliably across different hospitals and patient populations, rather than building a tool around the experience of a single institution"2

2

.

Clinical Implications and Future Directions

The study pursues two primary aims. First, researchers will develop and test models to predict whether a patient may develop postoperative acute kidney injury and assess potential severity. Second, they will identify patient groups that might benefit from different prevention strategies1

1

. Researchers will compare the federated learning models with existing prediction approaches and with a model built using pooled data from participating institutions. They will also assess performance across different patient populations, including groups that may be underrepresented in existing models. The findings are intended to inform future clinical trials that will test whether personalized prevention strategies can reduce postoperative acute kidney injury in patients undergoing surgery, potentially transforming how surgical teams approach kidney protection in the perioperative period2

2

.