A Real-Time Detection Algorithm for Electrocardiogram and Arterial Blood Pressure Signals in Intra-Aortic Balloon Pump

This paper proposes an algorithm combining a dynamic adaptive thresholding R-wave detection strategy and a differential feature-based blood pressure waveform recognition technique to realize rapid and real-time electrocardiogram (ECG) and arterial blood pressure (ABP) signal detection. The aim is to address the demand for accuracy and real-time performance of the intra-aortic balloon pump (IABP) system in heart failure treatment as an adjunct to heart failure. The proposed algorithm was validated in two public databases: the MIT-BIH arrhythmia database (MITDB) and the polysomnography database (SLPDB). The results showed that the sensitivity and positive predictive value of the ECG R-wave detection were 97.51 % and 99.73%, respectively, while the accuracy of the blood pressure feature point detection could reach 99.18%. Furthermore, the real-time performance of the algorithm was also evaluated, and the algorithm's detection average lead time was +0.46 milliseconds. In summary, this algorithm provides effective help for IABP timing technology and shows good application potential in the field of mechanical circulation-assisted therapy of cardiovascular diseases.