Semantic Segmentation of 12-Lead ECG Using 1D Residual U-Net with Squeeze-Excitation Blocks

Article Computer Science, Information Systems

FM-ECG: A fine-grained multi-label framework for ECG image classification

Nan Du et al.

Summary: The FM-ECG framework addresses the challenges of detecting abnormalities in real-world clinical ECG data by directly detecting abnormalities on ECG images with weakly supervised fine-grained classification and considering ECG label dependencies, outperforming other state-of-the-art methods.

INFORMATION SCIENCES (2021)

添加到收藏夹

Article Multidisciplinary Sciences

Transfer learning for ECG classification

Kuba Weimann et al.

Summary: Remote monitoring devices enable more effective healthcare for patients with periodic heart arrhythmia, but there is a need to develop reliable automatic ECG interpretation methods. This study uses deep learning to classify raw ECG recordings and shows that transfer learning improves classification performance effectively.

SCIENTIFIC REPORTS (2021)

添加到收藏夹

Article Engineering, Mechanical

1D convolutional neural networks and applications: A survey

Serkan Kiranyaz et al.

Summary: CNNs have become the standard for various Computer Vision and Machine Learning operations, with 1D CNNs offering advantages in applications with 1D signals. This paper provides a comprehensive review of 1D CNNs architecture, principles, and their recent progress in engineering applications.

MECHANICAL SYSTEMS AND SIGNAL PROCESSING (2021)

添加到收藏夹

Article Computer Science, Information Systems

Automatic Detection of QRS Complexes Using Dual Channels Based on U-Net and Bidirectional Long Short-Term Memory

Runnan He et al.

Summary: A novel algorithm based on U-Net and bidirectional long short-term memory with dual channels was proposed for automatic detection of QRS complexes in ECG signals, achieving high accuracy and sensitivity on two datasets after training and testing.

IEEE JOURNAL OF BIOMEDICAL AND HEALTH INFORMATICS (2021)

添加到收藏夹

Article Biology

ECGAug: A novel method of generating augmented annotated electrocardiogram QRST complexes and rhythm strips

Hans Friedrich Stabenau et al.

Summary: The application of neural networks in medicine has seen a significant rise, including in ECG segmentation. Existing methods for time series data augmentation may produce non-physiological signals and lack accurate fiducial point locations. ECGAug is a new method developed to generate an augmented training set with accurate fiducial point annotations, significantly improving neural network performance in training datasets.

COMPUTERS IN BIOLOGY AND MEDICINE (2021)

添加到收藏夹

Review Biology