تجاوز إلى المحتوى الرئيسي
User Image

Ridha OUNI

Associate Professor

Faculty

علوم الحاسب والمعلومات
CCIS, Building 31, Office No 2193, P. O. Box 51178, Riyadh 11543
المنشورات
مقال فى مجلة
2020

Electrocardiogram heartbeat classification based on a deep convolutional neural network and focal loss, Computers in Biology and Medicine

Taissir Fekih Romdhanea, Haikel Alhichri, Ridha Ouni, Mohamed Atri

The electrocardiogram (ECG) is an effective tool for cardiovascular disease diagnosis and arrhythmia detection. Most methods proposed in the literature include the following steps: 1) denoizing, 2) segmentation into heartbeats, 3) feature extraction, and 4) classification. In this paper, we present a deep learning method based on a convolutional neural network (CNN) model. CNN models can perform feature extraction automatically and jointly with the classification step. In other words, our proposed method does not require a feature extraction step with hand-crafted techniques. Our proposed method is also based on an algorithm for heartbeat segmentation that is different from most existing methods. In particular, the segmentation algorithm defines each ECG heartbeat to start at an R-peak and end after 1.2 times the median RR time interval in a 10-s window. This method is simple and effective, as it does not use any form of filtering or processing that requires assumptions about the signal morphology or spectrum. Although enhanced ECG heartbeat classification algorithms have been proposed in the literature, they failed to achieve high performance in detecting some heartbeat categories, especially for imbalanced datasets. To overcome this challenge, we propose an optimization step for the deep CNN model using a novel loss function called focal loss. This function focuses on minority heartbeat classes by increasing their importance. We trained and evaluated our proposed model with the MIT-BIH and INCART datasets to identify five arrhythmia categories (N, S, V, Q, and F) based on the Association for Advancement of Medical Instrumentation (AAMI) standard. The evaluation results revealed that the focal loss function improved the classification accuracy for the minority classes as well as the overall metrics. Our proposed method achieved 98.41% overall accuracy, 98.38% overall F1-score, 98.37% overall precision, and 98.41% overall recall. In addition, our method achieved better performance than that of existing state-of-the-art methods.

نوع عمل المنشور
Ph.D
رقم المجلد
123
مجلة/صحيفة
Computers in Biology and Medicine
الصفحات
pp. 1-13
مزيد من المنشورات
publications
بواسطة Sulaiman Alwabisi, Ridha Ouni, and Kashif Saleem
2022
تم النشر فى:
Electronics
publications
بواسطة 2. A. Alashban, A. Alsadan, N. F. Alhussainan and R. Ouni
2022
تم النشر فى:
IEEE access
publications
بواسطة Ridha Ouni, Kashif Saleem
2022
تم النشر فى:
Sustainability