Search

The objective of this article is to develop methods and algorithms for detecting QRS complexes and cardiac episodes, primarily diagnosing tachy- and bradyarrhythmias, as well as heart pauses and atrial fibrillation. Algorithms and methods use parameters such as ECG, physical activity and the patient's body position. Datasets were collected by BioModule BioHarness 3 (Zephyr Technology Corporation, Annapolis, MD). The monitor was installed simultaneously with Holter monitor MYOCARD-HOLTER (NIMP ESN LLC, Sarov, Nizhny Novgorod region). Data from the PhysioNet database was used to verify the QRS detector and RR-interval analyzer. The algorithm is validated using the PhysioNet database, resulting a sensitivity of Se = 99.18% and a specificity Sp = 99.52%, and using the BioMonitor 3, resulting a sensitivity of Se = 98.96% and a specificity Sp = 93.37%. 20 records were analyzed for the presence event such as tachycardia and bradycardia, cardiac pauses and atrial fibrillation. 18 of 20 were recognized correctly by the analyzer. The analyzer gave an incorrect result in 2 cases. [ABSTRACT FROM AUTHOR]