Date of Award
Master of Science and Software Engineering (MSSE)
Mohammad Mehedy Masud
M. Adel Serhani
Cardiovascular arrhythmia diseases are considered as the most common diseases that cause death around the world. Abnormal arrhythmia diseases can be identified by analyzing heart rhythm using an electrocardiogram (ECG). However, this analysis is done manually by cardiologists, which may be subjective and susceptible to different cardiologist observations and experiences, as well as to noise and irregularities in those signals. This can lead to misdiagnosis. Motivated by this challenge, an automated heart rhythm diagnosis approach from ECG signals using Deep Learning has been proposed. In order to achieve this goal, three research problems have been addressed. First, recognize the role of each single-lead of a 12-lead ECG to classify heart rhythms. Second, understanding the importance of static data (e.g., demographics and clinical profile) in classifying heart rhythms. Third, realizing whether the static data can be combined with the ECG time series data for better classification performance. In this thesis, different deep learning models have been proposed to address these problems and satisfactory results are achieved. Therefore, using this knowledge, an effective hybrid deep learning model to classify heart rhythms has been proposed. As per knowledge obtained from relevant literature, this is the first work to identify the importance of individual lead and combined lead as well as the importance of combining static data with ECG time series data in classifying heart rhythms. Extensive experiments have been performed to evaluate this algorithm on a 12-lead ECG database that contains data from more than 10,000 individual subjects and obtained a high average of accuracy (up to 98.7%) and F1-measure (up to 98.7%). Moreover, in this thesis, the distribution of heart rhythms from the database based on heart rhythm type, gender, and age group have been analyzed, which will be valuable for further improvement of classification performance. This study will provide valuable insights and will prove to be an effective tool in automated heart rhythm classification and will assist cardiologists in effectively and accurately diagnosing heart disease.
Alnajjar, Ahmad Abdulrazaq Abdulla, "HEART RHYTHM CLASSIFICATION FROM STATIC AND ECG TIME-SERIES DATA USING HYBRID MULTIMODAL DEEP LEARNING" (2021). Theses. 844.