A Multi-Level Classification Approach for Sleep Stage Prediction With Processed Data Derived From Consumer Wearable Activity Trackers.
A Multi-Level Classification Approach for Sleep Stage Prediction With Processed Data Derived From Consumer Wearable Activity Trackers.
复制标题
DOI：
10.3389/fdgth.2021.665946
复制
发表时间：
2021
影响因子：
--
通讯作者：
Chapa-Martell MA
中科院分区：
文献类型：
--
作者：
Liang Z;Chapa-Martell MA
消费者可穿戴活动跟踪器，如Fitbit，广泛用于自由生活环境中无处不在的纵向睡眠监测。然而，已知这些设备对于测量睡眠阶段是不准确的。在这项研究中，我们开发并验证了一种新的方法，该方法利用了消费者活动跟踪器（即，步数、心率和睡眠指标）来预测睡眠阶段。该方法采用选择性校正策略，由两级分类器组成。I级分类器判断Fitbit标记的睡眠时期是否被错误分类，并且II级分类器将错误分类的时期重新分类为四个睡眠阶段之一（即，浅睡眠、深睡眠、REM睡眠和觉醒）。当支持向量机和梯度提升决策树（XGBoost）与上采样时，分别在I级和II级分类时，实现了最佳的逐时性能。该模型实现了0.731 ± 0.119的总体每历元准确度，0.433 ± 0.212的Cohen's Kappa，以及0.451 ± 0.214的多类Matthew's相关系数（MMCC）。关于个体睡眠阶段的总持续时间，该模型的平均归一化绝对偏差（MAB）为0.469，与专有Fitbit算法相比降低了23.9%。将支持向量机和XGBoost与下采样相结合的模型实现了0.704 ± 0.097的次优每历元精度，Cohen's Kappa为0.427 ± 0.178，MMCC为0.439 ± 0.180。次优模型获得的MAB为0.179，与专有Fitbit算法相比显著降低了71.0%。我们强调了基于机器学习的睡眠阶段预测与消费者可穿戴设备的挑战，并为未来的研究提出了方向。
Consumer wearable activity trackers, such as Fitbit are widely used in ubiquitous and longitudinal sleep monitoring in free-living environments. However, these devices are known to be inaccurate for measuring sleep stages. In this study, we develop and validate a novel approach that leverages the processed data readily available from consumer activity trackers (i.e., steps, heart rate, and sleep metrics) to predict sleep stages. The proposed approach adopts a selective correction strategy and consists of two levels of classifiers. The level-I classifier judges whether a Fitbit labeled sleep epoch is misclassified, and the level-II classifier re-classifies misclassified epochs into one of the four sleep stages (i.e., light sleep, deep sleep, REM sleep, and wakefulness). Best epoch-wise performance was achieved when support vector machine and gradient boosting decision tree (XGBoost) with up sampling were used, respectively at the level-I and level-II classification. The model achieved an overall per-epoch accuracy of 0.731 ± 0.119, Cohen's Kappa of 0.433 ± 0.212, and multi-class Matthew's correlation coefficient (MMCC) of 0.451 ± 0.214. Regarding the total duration of individual sleep stage, the mean normalized absolute bias (MAB) of this model was 0.469, which is a 23.9% reduction against the proprietary Fitbit algorithm. The model that combines support vector machine and XGBoost with down sampling achieved sub-optimal per-epoch accuracy of 0.704 ± 0.097, Cohen's Kappa of 0.427 ± 0.178, and MMCC of 0.439 ± 0.180. The sub-optimal model obtained a MAB of 0.179, a significantly reduction of 71.0% compared to the proprietary Fitbit algorithm. We highlight the challenges in machine learning based sleep stage prediction with consumer wearables, and suggest directions for future research.
登录
查看更多内容
影响因子：
4.3
作者：
Finan, Patrick H.;Richards, Jessica M.;Smith, Michael T.
通讯作者：
Smith, Michael T.
影响因子：
4.8
作者：
Fisher, RA;Corbet, AS;Williams, CB
通讯作者：
Williams, CB
影响因子：
2.9
作者：
de Zambotti, Massimiliano;Baker, Fiona C.;Colrain, Ian M.
通讯作者：
Colrain, Ian M.
影响因子：
3.7
作者：
Delgado, Rosario;Tibau, Xavier-Andoni
通讯作者：
Tibau, Xavier-Andoni
影响因子：
2.7
作者：
FLEISS, JL;COHEN, J
通讯作者：
COHEN, J