A Proficient Evalution with the Pre-Term Birth Classification in ECG Signal using KNN

The rate of premature births is rising all over the world and there is still no forecast against the births. Recent research is based on ECG record analysis, which contains information on the electrophysiological properties of the mother's and fetal's cardiac signals. The purpose of this work is to classify the fetal ECG heartbeats, using the KNN classifier, and to predict preterm birth. 50 ECG signals were taken in this work, it is preprocessed using the FIR and NLMS filters. Using FFT the function was extracted based on the pre-processed signals. Classification of the signals with the extracted features is unclear. The features were optimized The functions were designed using PSO and chosen. In addition, identify the ECG signals according to the qualified features, selected features and target value. ECG signals were marked as a term or premature with 93.4045 percent accuracy, 99.596 percent sensitivity and 88.3268 percent specificity.