Identifying artifacts in arterial blood pressure using morphogram variability

In this article we present a simple technique that utilizes the cross correlations between ECG signals and haemodynamic signals for the purpose of assessing signal quality and detecting artifacts in the arterial blood pressure (ABP) signal. The technique was tested using cases from a physician-annotated patient monitoring signal database from Beth IsraeUHarvard-MIT University data bank. The results were encouraging: 90% of the manually annotated artifacts were correctly classified as artifacts and 99% of the manually annotated true events were correctly classified (out of a tQtd Of 683 manualiy annotated alarms).