Implementation of a Simple Real-Time Algorithm for Ventricular Fibrillation Detection in a Microcontroller

This paper describes the implementation of an algorithm for detecting episodes of Ventricular Fibrillation (VF) in a MSP430F6659 microcontroller. Statistical and digital signal processing techniques were used in its design. The validity of its performance was carried out in 3 ECG records from the MIT-BIH Arrhythmia Database, 1 record of the MIT-BIH Malignant Ventricular Arrhythmia Database and in 10 records of the Creighton University Ventricular Tachyarrhythmia Database. These records correspond to patients who suffered of VF episodes during the ECG study. The algorithm was able to detect 100 % of VF events and has a time of computation of approximately 100 μs.