A computational model for heart failure stratification

Heart failure (HF) is a kind of serious cardiovascular diseases, leading to an increasing burden imposed on public healthcare. Early diagnosis and proper treatment of HF are essential to reducing its morbidity and mortality. In spite of the implementation of clinical guidelines for HF, early recognition and stratification of HF risk remain unsolved. In this work, we supposed a computational model to classify HF stages. With aid of Monte Carlo simulation, Naïve Bayesian Classifier (NBC), Support Vector Machine (SVM) and Radial Basis Function Network (RBF) etc. models were investigated. On the basis of the model assessment, an optimum classification model constitutive of SVM was derived. The model was tested on 389 subjects. The results show that 81.06% cases in total are consistent with the outcomes by AHA/ACC staging system. This work may offer a quantitative tool for HF stratification and facilitate early diagnosis for HF.