Continuous long-term monitoring of electrocardiography (ECG) signals is
crucial for the early detection of cardiac abnormalities such as arrhythmia.
Non-clinical ECG recordings acquired by Holter and wearable ECG sensors often
suffer from severe artifacts such as baseline wander, signal cuts, motion
artifacts, variations on QRS amplitude, noise, and other interferences.
Usually, a set of such artifacts occur on the same ECG signal with varying
severity and duration, and this makes an accurate diagnosis by machines or
medical doctors extremely difficult. Despite numerous studies that have
attempted ECG denoising, they naturally fail to restore the actual ECG signal
corrupted with such artifacts due to their simple and naive noise model. In
this study, we propose a novel approach for blind ECG restoration using
cycle-consistent generative adversarial networks (Cycle-GANs) where the quality
of the signal can be improved to a clinical level ECG regardless of the type
and severity of the artifacts corrupting the signal. To further boost the
restoration performance, we propose 1D operational Cycle-GANs with the
generative neuron model. The proposed approach has been evaluated extensively
using one of the largest benchmark ECG datasets from the China Physiological
Signal Challenge (CPSC-2020) with more than one million beats. Besides the
quantitative and qualitative evaluations, a group of cardiologists performed
medical evaluations to validate the quality and usability of the restored ECG,
especially for an accurate arrhythmia diagnosis.

本文提出了一个新方法，使用 Cycle-GAN 来盲目去噪 ECG 信号，从而提高信号质量，无论信号受到何种类型和严重程度的干扰，使其恢复成临床水平的 ECG。该方法在 CPSC-2020 ECG 数据集上经过了量化和定性评估，并得到了心脏病专家的验证。