Is a higher sampling rate desirable in the computer processing of the pediatric electrocardiogram?

Abstract

A considerably large number of ECGs of pediatric patients had small notches and slurs on QRS complexes. By sampling these ECGs at higher rates, namely 4 KHz, it was shown that the higher rate data accurately reconstructed the notches and slurs on the QRS complex which were occasionally missed or distorted when the sampling rate was 250 Hz. When using the sampling rate of 250 Hz, it was observed that the reduction of the peak-to-peak amplitude of a QRS complex exceeded 0.1 mV in nearly 30 percent of tested cases, even if the ECG did not contain unwanted noise signals. Moreover, if there is relatively more noise in children's ECGs, one should reduce noise in such data in order to prevent measurement errors in the succeeding steps of the computer program. In such circumstances, the higher sampling rate would be preferable since filtering techniques which are used to reduce noise may considerably affect the estimation of the onset, offset and amplitude of the intrinsic ECG signals if the sampling rate is low. The reduction of the price of computers and their peripheral devices may facilitate the adoption of the higher sampling rate. From these considerations it was concluded that the use of the higher sampling rate might be both practical and an effective measure to enhance the diagnostic performance of the computer ECG systems.