Fast non-invasive ventricular fibrillation detection method using pseudo Wigner-Ville distribution

Detection of ventricular fibrillation (VF) at an early stage is crucial in order to lower the risk of sudden death and allow the specialist to have a greater reaction time to give the patient a good recovering therapy and avoid unrecoverable damage to the cardiac tissue. We present an algorithm oriented towards the real-time detection of VF, to be used as a part of monitoring systems in intensive care units or ambulatory centres. The study has been done using the AHA (American Heart Association) database, focusing mainly on the 8200 series, and the MIT (Massachussetts Institute of Technology) database. The detection algorithm combines both time-domain and time-frequency parameters. Using the appropriate parameters, the detection algorithm discerns between VF and non-VF rhythms, including VF-like rhythms like certain types of ventricular tachycardia (VT). A VF flutter sensitivity of 86% and an average specificity of 94.3 % (including VT separation) is obtained