University of Twente Student Theses
Using lightweight image classifiers for electrocardiogram classification on embedded devices
Heinen, Niek (2020) Using lightweight image classifiers for electrocardiogram classification on embedded devices.
|
PDF
651kB |
| Abstract: | A large part of deep learning research is devoted to image classification. The research in this paper will show that the same neural networks developed for image classification can also be used to accurately classify electrocardiograms (ECGs). Even though this is a novel approach for ECG classification, the early results appear to be promising. Since both accuracy and efficiency are much valued in the applications of ECG classifiers the research will focus on using lightweight image classifiers such as ResNet and MobileNet. In order to (maximally) utilise the architecture of the image classifiers, we need to cleverly reshape the ECG signals. In this paper, we will investigate numerous ways of doing so. Using the VEB en SVEB evaluation metrics, the research in this paper will be compared to the state-of-the-art. The best performing ECG classifier presented in this paper achieved a VEB- and SVEB-accuracy of 96.8% and 98.1% respectively. |
| Item Type: | Essay (Bachelor) |
| Faculty: | EEMCS: Electrical Engineering, Mathematics and Computer Science |
| Subject: | 54 computer science |
| Programme: | Computer Science BSc (56964) |
| Link to this item: | https://purl.utwente.nl/essays/82240 |
| 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