Reducing ECG alarm fatigue based on SQI analysis

Alarm fatigue can cause many negative results. Regarding ECG signals, an algorithm to reduce alarm fatigue is described, regardless of whether or not VPBs are involved in the false alarms. ECG signals are divided into five signal quality index levels (SQI): 0∼4, where level 0 represents a noise free signal and level 4 indicates the worst signal quality. The key of the method is to judge the noise level by recognizing P, QRS, and T waves effectively. If the SQI value is found to be high, some strategies can be implemented to reduce alarm fatigue, including improvement of the classification of VPB-like QRS complexes, correction of some false ARR alarms, and freeze of the HR value for a few seconds. The AHA and MIT-BIH database as well as the Mindray database were used to evaluate the performance of the SQI analysis. Compared with the new method, QRS and VPB detection accuracies are almost the same based on DB1, and both incorrect HR values and false ARR alarms can reduced more than 40% based on DB2, DB3, and DB4.