Combining ECG analysis with clinical parameters for diagnosis of heart failure

Congestive heart failure (CHF) is a continuing health problem. The purpose of the work described here is to develop a computer-assisted decision model which can aid in the differentiation of CHF from other types of cardiovascular disease. Two methods developed by the authors are combined to produce this model. The first is the use of continuous chaotic modeling to analyze 24-hour Holter data. The second is a neural network model which combines the results of the Holter analysis with clinical parameters to establish the final model. Preliminary results indicate that the combined model offers a promising approach for differentiation of CHF patients from patients with other cardiovascular disease.