当前位置: X-MOL 学术Int. J. Neural Syst. › 论文详情
Our official English website, www.x-mol.net, welcomes your feedback! (Note: you will need to create a separate account there.)
Time–Frequency Decomposition of Scalp Electroencephalograms Improves Deep Learning-Based Epilepsy Diagnosis
International Journal of Neural Systems ( IF 6.6 ) Pub Date : 2021-07-16 , DOI: 10.1142/s0129065721500325
Prasanth Thangavel 1 , John Thomas 1 , Wei Yan Peh 1 , Jin Jing 2 , Rajamanickam Yuvaraj 1, 3 , Sydney S Cash 2 , Rima Chaudhari 4 , Sagar Karia 5 , Rahul Rathakrishnan 6 , Vinay Saini 7 , Nilesh Shah 5 , Rohit Srivastava 7 , Yee-Leng Tan 8 , Brandon Westover 2 , Justin Dauwels 1, 9
Affiliation  

Epilepsy diagnosis based on Interictal Epileptiform Discharges (IEDs) in scalp electroencephalograms (EEGs) is laborious and often subjective. Therefore, it is necessary to build an effective IED detector and an automatic method to classify IED-free versus IED EEGs. In this study, we evaluate features that may provide reliable IED detection and EEG classification. Specifically, we investigate the IED detector based on convolutional neural network (ConvNet) with different input features (temporal, spectral, and wavelet features). We explore different ConvNet architectures and types, including 1D (one-dimensional) ConvNet, 2D (two-dimensional) ConvNet, and noise injection at various layers. We evaluate the EEG classification performance on five independent datasets. The 1D ConvNet with preprocessed full-frequency EEG signal and frequency bands (delta, theta, alpha, beta) with Gaussian additive noise at the output layer achieved the best IED detection results with a false detection rate of 0.23/min at 90% sensitivity. The EEG classification system obtained a mean EEG classification Leave-One-Institution-Out (LOIO) cross-validation (CV) balanced accuracy (BAC) of 78.1% (area under the curve (AUC) of 0.839) and Leave-One-Subject-Out (LOSO) CV BAC of 79.5% (AUC of 0.856). Since the proposed classification system only takes a few seconds to analyze a 30-min routine EEG, it may help in reducing the human effort required for epilepsy diagnosis.

中文翻译:

头皮脑电图的时频分解改善了基于深度学习的癫痫诊断

基于头皮脑电图 (EEG) 中发作间期癫痫样放电 (IED) 的癫痫诊断费力且通常具有主观性。因此,有必要建立一种有效的 IED 检测器和一种自动分类无 IED 与 IED 脑电图的方法。在这项研究中,我们评估了可能提供可靠的 IED 检测和 EEG 分类的特征。具体来说,我们研究了基于具有不同输入特征(时间、光谱和小波特征)的卷积神经网络 (ConvNet) 的 IED 检测器。我们探索了不同的 ConvNet 架构和类型,包括 1D(一维)ConvNet、2D(二维)ConvNet,以及各个层的噪声注入。我们在五个独立的数据集上评估 EEG 分类性能。具有预处理的全频脑电信号和频带(delta、theta、alpha、beta)的 1D ConvNet 在输出层具有高斯加性噪声,在 90% 的灵敏度下实现了最佳的 IED 检测结果,误检率为 0.23/min。EEG 分类系统获得的平均 EEG 分类 Leave-One-Institution-Out (LOIO) 交叉验证 (CV) 平衡准确度 (BAC) 为 78.1%(曲线下面积 (AUC) 为 0.839)和 Leave-One-Subject -Out (LOSO) CV BAC 为 79.5%(AUC 为 0.856)。由于所提出的分类系统只需几秒钟即可分析 30 分钟的常规脑电图,它可能有助于减少癫痫诊断所需的人力。EEG 分类系统获得的平均 EEG 分类 Leave-One-Institution-Out (LOIO) 交叉验证 (CV) 平衡准确度 (BAC) 为 78.1%(曲线下面积 (AUC) 为 0.839)和 Leave-One-Subject -Out (LOSO) CV BAC 为 79.5%(AUC 为 0.856)。由于所提出的分类系统只需几秒钟即可分析 30 分钟的常规脑电图,它可能有助于减少癫痫诊断所需的人力。EEG 分类系统获得的平均 EEG 分类 Leave-One-Institution-Out (LOIO) 交叉验证 (CV) 平衡准确度 (BAC) 为 78.1%(曲线下面积 (AUC) 为 0.839)和 Leave-One-Subject -Out (LOSO) CV BAC 为 79.5%(AUC 为 0.856)。由于所提出的分类系统只需几秒钟即可分析 30 分钟的常规脑电图,它可能有助于减少癫痫诊断所需的人力。
更新日期:2021-07-16
down
wechat
bug