EEGdenoiseNet
收藏资源简介:
EEGdenoiseNet是由南方科技大学深圳智能医疗工程实验室创建的一个专为EEG去噪深度学习模型训练和测试设计的基准数据集。该数据集包含4514个纯净EEG片段、3400个眼电图(EOG)片段和5598个肌电图(EMG)片段,允许用户合成带有真实纯净EEG的污染EEG片段。数据集通过预处理和专家视觉检查确保数据质量,适用于评估和比较不同深度学习模型的去噪性能。EEGdenoiseNet的应用领域主要集中在解决EEG信号中的噪声和伪影问题,特别是在心理学、神经学和精神病学研究以及脑机接口中。
EEGdenoiseNet is a benchmark dataset created by the Shenzhen Intelligent Medical Engineering Laboratory of Southern University of Science and Technology, specifically designed for training and testing deep learning models for EEG denoising. This dataset contains 4514 clean EEG segments, 3400 electrooculogram (EOG) segments, and 5598 electromyogram (EMG) segments, enabling users to synthesize contaminated EEG segments paired with their corresponding ground-truth clean EEGs. The dataset ensures data quality through preprocessing and expert visual inspection, and is suitable for evaluating and comparing the denoising performance of various deep learning models. The application fields of EEGdenoiseNet mainly focus on addressing noise and artifact problems in EEG signals, particularly in psychological, neurological, and psychiatric research, as well as brain-computer interfaces.




