University of Twente Student Theses
Electrocardiographic analysis of ventricular fibrillation: a machine learning approach to identify a previous myocardial infarction
Gantevoort, S. (2018) Electrocardiographic analysis of ventricular fibrillation: a machine learning approach to identify a previous myocardial infarction.
|
PDF
11MB |
| Abstract: | Out-of-hospital cardiac arrest (OHCA) is one of the main causes of death and carries a poor prognosis. Ventricular fibrillation (VF) as initial rhythm is frequently observed in OHCA cases. Quantitative measures of the VF waveform have been investigated to optimize resuscitation strategies in the OHCA setting. It has already been demonstrated that these VF characteristics are related to arrest duration and shock success. Research has shown that myocardial infarction (MI) influences the VF waveform as well. MI, defined as myocardial cell necrosis due to prolonged ischemia, is the most common cause of VF and can only be diagnosed after restored circulation. Notably, an MI is commonly reversible, but cannot be determined in-field. As MI affects the VF waveform, it might be possible to identify MI using the VF waveform, potentially enabling patient tailored treatment of OHCA patients. Previous research has aimed to detect a previous MI in a controlled setting. In a proof of concept study, the potential of machine learning algorithms to identify an MI using a 12-lead electrocardiogram (ECG) seemed feasible. However, translation to the OHCA setting is limited, because in the acute setting only a single ECG-lead is measured by the defibrillator paddles. In light of the above, we investigated the ability of a single ECG-lead to differentiate between patient with and without a history of MI, in a controlled setting of implantable cardioverter defibrillator (ICD) implantations. In addition, we investigated whether multiple ECG-leads may be superior to the single ECG-lead approach. In follow-up on this analysis, an optimal lead combination for the multiple lead approach was investigated. |
| Item Type: | Essay (Master) |
| Clients: | Radboud University Medical Centre, Nijmegen, The Netherlands |
| Faculty: | TNW: Science and Technology |
| Subject: | 44 medicine, 50 technical science in general |
| Programme: | Technical Medicine MSc (60033) |
| Link to this item: | https://purl.utwente.nl/essays/76225 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page