AI 中文总结
本研究提出基于学习的个性化评分者建模框架 LBH，利用多评分者数据集构建更可靠的睡眠阶段参考标签，在 DOD-H、DOD-O 数据集上较基线方法提升了睡眠分期性能。
AI 中文摘要
睡眠阶段分类对睡眠障碍的诊断和管理至关重要，但大多数自动分期研究在评估模型时仅采用单一参考 hypnogram（ hypnogram 即 hypnogram，睡眠图），尽管已知评分者间存在变异性。本研究探究是否可利用多评分者数据集，从多位专家的集体行为中构建更可靠的参考标签。研究使用公开可用的 DOD-H 和 DOD-O 数据集，将 EEG（C3-M2）和颏肌电图（chin EMG）信号分割为 30 秒的 epoch（epoch 即 epoch，时间窗），并从每种模态中提取 30 个特征，EEG+EMG 模态共得到 60 个特征。我们提出一种基于学习的睡眠图（LBH），该方法利用机器学习模型得到的混淆矩阵，对每位评分者的阶段特定行为进行建模；经列归一化后，这些矩阵可估计每个真实睡眠阶段对应每位评分者标签的概率，再将多位评分者的概率聚合，为每个时间窗分配最终标签。LBH 在仅用 EEG 和 EEG+EMG 两种设置下，分别采用随机森林、支持向量机、多层感知机分类器进行评估，并与数据集睡眠图（DH）和最佳评分者睡眠图（BSH）对比。LBH 在所有情况下均提升了整体性能，在随机森林与 EEG+EMG 的组合下取得最佳结果：在 DOD-H 数据集上达到 86.07% 的准确率、85.46% 的精确率和 85.29% 的 F1 值，在 DOD-O 数据集上达到 86.04% 的准确率、85.21% 的精确率和 84.70% 的 F1 值。这些结果表明，个性化评分者建模可在不丢弃单个专家信息的前提下，提升参考睡眠图的构建质量。
英文摘要
Sleep stage classification is important for the diagnosis and management of sleep disorders, yet most automatic staging studies evaluate models against a single reference hypnogram despite known inter-scorer variability. This study investigates whether multi-scored datasets can be used to construct more reliable reference labels from the collective behavior of multiple experts. We use the publicly available DOD-H and DOD-O datasets. EEG (C3-M2) and chin EMG signals were segmented into 30-s epochs, and 30 features were extracted from each modality, yielding 60 features for EEG+EMG. We propose a learning-based hypnogram (LBH) that models the stage-specific behavior of each scorer using confusion matrices derived from machine-learning models. After column normalization, these matrices estimate the probability of each true sleep stage given each scorer's label; probabilities are aggregated across scorers to assign the final label for each epoch. LBH was evaluated with random forest, support vector machine, and multilayer perceptron classifiers under EEG-only and EEG+EMG settings, and compared with the dataset hypnogram (DH) and best-scorer hypnogram (BSH). LBH consistently improved overall performance. The best results were obtained with random forest and EEG+EMG, reaching 86.07% accuracy, 85.46% precision, and 85.29% F1-score on DOD-H, and 86.04% accuracy, 85.21% precision, and 84.70% F1-score on DOD-O. These findings suggest that personalized scorer modeling can improve reference hypnogram construction without discarding information from individual experts.
