Machine learning-based risk prediction model for canine myxomatous mitral valve disease using electronic health record data

Kim, Y., Kim, J., Kim, S., Youn, H., Choi, J., & Seo, K. (2023). Machine learning-based risk prediction model for canine myxomatous mitral valve disease using electronic health record data. Frontiers in Veterinary Science, 10. https://doi.org/10.3389/fvets.2023.1189157

This study developed machine learning models to predict heart failure risk in dogs with myxomatous mitral valve disease (MMVD) using electronic health records. Data from 143 dogs between 2018 and 2022 were analyzed, incorporating demographic, radiographic, echocardiographic, and laboratory values. Four machine learning algorithms were tested, with the random forest model achieving the highest performance (AUC = 0.88), while K-nearest neighbors showed the lowest (AUC = 0.69). Echocardiographic and radiographic variables, along with chloride levels, were key predictors of heart failure.