Researchers at UVA Health have created a new risk assessment tool using machine learning and artificial intelligence to predict outcomes for heart failure patients. This tool is now available for free to clinicians and aims to improve care by identifying specific risks for each patient.Heart failure is a condition that affects both the quality and quantity of life. According to Dr. Sula Mazimba, a heart failure expert, it’s important to recognize that not all patients are the same and each one falls on a spectrum of risk for adverse outcomes. By identifying the level of risk for each patient, clinicians can personalize treatments to improve their outcomes.
Heart failure occurs when the heart cannot pump enough blood to meet the body’s needs, leading to symptoms like fatigue, weakness, and swollen legs and feet. Ultimately, it can result in death. Heart failure is a progressive condition.different heart failure studies and registries. The data included information on demographics, medical history, medications, lab results, and other clinical variables. The researchers then used this data to create a predictive model that could identify patients at high risk for adverse outcomes, such as hospitalization or death. This information is crucial for clinicians in order to provide the best care for their patients. Additionally, with the increasing prevalence of heart failure, the need for improved care is becoming more pressing. The researchers at UVA have developed the CARNA model to address this need and improve care for heart failure patients. This model was developed using data from thousands of patients in various studies and registries, and it aims to help identify high-risk patients for better management of their condition.Heart failure clinical trials that were previously supported by the National Institutes of Health’s National Heart, Lung and Blood Institute have found that a newly developed model has outperformed existing predictors in determining the outcomes for patients. The model was tested and found to be more effective in predicting the need for heart surgery or transplant, the risk of rehospitalization, and the risk of death for a broad spectrum of patients. The researchers attribute the model’s success to the use of ML/AI and its incorporation of “hemodynamic” clinical data, which provides information on how blood circulates through the heart, lungs, and the rest of the body. This innovative approach presents promising advancements in the field.University of Virginia School of Engineering’s Department of Computer Science has developed a breakthrough model that can process complex data and make decisions in the presence of missing or conflicting factors. According to researcher Josephine Lamp, this model is an exciting development because it intelligently presents and summarizes risk factors, making it easier for clinicians to make treatment decisions quickly.
The use of this model could potentially enable doctors to personalize care for their patients, ultimately leading to longer and healthier lives. The collaborative research environment at the University of Virginia is contributing to these important advancements in the field.The collaboration of experts in heart failure, computer science, data science, and statistics, led by researcher Kenneth Bilchick, MD, a cardiologist at UVA Health, has made this work in Virginia possible. According to Bilchick, “Multidisciplinary biomedical research that integrates talented computer scientists like Josephine Lamp with experts in clinical medicine will be crucial in helping our patients benefit from AI in the years to come.”
The researchers have published their findings and made their new tool available online for free at https://github.com/jozieLamp/CARNA.The findings of their evaluation of CARNA in the American Heart Journal have been published by the research team, which included Lamp, Yuxin Wu, Steven Lamp, Prince Afriyie, Nicholas Ashur, Bilchick, Khadijah Breathett, Younghoon Kwon, Song Li, Nishaki Mehta, Edward Rojas Pena, Lu Feng, and Mazimba. The researchers have no financial interest in the project.
This project was based on a winning submission to the National Heart, Lung and Blood Institute’s Big Data Analysis Challenge: Creating New Paradigms for Heart Failure Research. It was supported by the National Science Foundation Graduate Research Fellowship, grant 8424.90, and NHLBI grants R56HL159216, K01HL142848 and L30HL148881.
