Bibliographic Review of methods of detection of Ventricular Fibrillation based on ECG signals

Ventricular fibrillation is one of the most dangerous arrhythmias, because it causes a chaotic heart rhythm that can lead to cardiac arrest leading to sudden death in individuals. For this reason, detecting this disease on time gives the specialist doctor the possibility of treating it and increases the life expectancy of the patient. The objective of this research work is to compile the different ventricular fibrillation detection methods that can be obtained from the ECG signal from databases such as: IEEEexplore, ScienceDirect, Scopus, etc., emphasizing the work of the last 10 years, based on the relevance of the topic, the credentials of the authors and objectivity. Ventricular fibrillation detection methods are based on characteristics or patterns found in ECG signals that allow them to be recognized among the signals of other arrhythmia. However, the methods have evolved and are combined with artificial intelligence algorithms such as neural networks and machine learning techniques to improve the detection of VF. The results are evaluated using databases that cover different age ranges and hospital and extra-hospital patients, such as the MIT-BIH arrhythmia and ventricular arrhythmia databases and the Creigton University Ventricular Tachyarrhymia database; and the parameters that validate these results are the sensitivity, specificity and precision. Among the best performing methods for ventricular fibrillation detection are algorithms that feature preprocessing of the ECG signal in which much of the noise is removed. This facilitates the phase of searching for characteristics or patterns within the signal ECG and finally the detection of ventricular fibrillation based on machine learning methods with an average sensitivity and specificity between 80% and 90%.

• Hovedforfatter: Tacuri Pineda, José Andrés (author)
• Format: bachelorThesis
• Sprog: eng
• Udgivet: 2020
• Fag: Ventricular Fibrillation; Detection; Methods; ECG; Sensitivity; Precision; Artificial intelligence; Algorithms; Fibrilación ventricular; Detección; Métodos; Sensibilidad; Precisión; Inteligencia artificial; Algoritmos
• Online adgang: http://repositorio.yachaytech.edu.ec/handle/123456789/280