An ECG signal processor is used to predict cardiac arrest which results in low power consumption.unique set of ECG features and Naive Bayes classifier is used in this design. The objective is to reduce power consumption by using clock gating. The heart signal recordings are taken from MIT-PhysioNet. To detect the QRS complex wave Soft threshold algorithm is used to avoid distortion in the ECG signal. This system of design consist of three stages includes preprocessing, feature extraction and classification of the ECG signal .The Naive Bayes Classifier is used to compare thesignals which is taken from the MIT-Physionet data basae. The fiducial points of T and P wave are taken from the ECG signal.