Haghayegh Shahab, Khoshnevis Sepideh, Smolensky Michael H, Diller Kenneth R
Department of Biomedical Engineering Cockrell School of Engineering, The University of Texas at Austin, USA.
Department of Biomedical Engineering Cockrell School of Engineering, The University of Texas at Austin, USA.
Sleep Med. 2020 Oct;74:235-241. doi: 10.1016/j.sleep.2020.05.008. Epub 2020 May 15.
Estimation of sleep parameters by wrist actigraphy is highly dependent on performance of the interpretative algorithm (IA) that converts movement data into sleep/wake scores.
(1) Does the actigraphy mode of operation -Proportional Integrating Measure (PIM) or Zero Crossing Mode (ZCM), responsive respectively to intensity and frequency of movements- impact sleep scoring; and (2) Can a better performing sleep scoring IA be developed by a deep learning approach combining PIM/ZCM data.
ZCM and PIM plus electroencephalographic (EEG) data of 40 healthy adults (17 female, mean age: 26.7 years) were obtained from a single in-home nighttime sleep study. Effect of mode of operation was first evaluated by applying several classic deep learning models to PIM only, ZCM only, and combined ZCM/PIM data. After, a novel deep learning model was developed incorporating combined ZCM/PIM data, and its performance was compared with existing Cole-Kripke, rescored Cole-Kripke, Sadeh, and UCSD IAs.
Relative to the EEG reference, ZCM/PIM combined mode produced higher agreement of scoring sleep/wake epochs than only ZCM or PIM modes. The proposed novel deep learning model showed 87.7% accuracy (0.2-1% higher than the other IAs), 94.1% sensitivity (0.7-4.3% lower than the other IAs), 64.0% specificity (9.9-21.5% higher than the other IAs), and 59.9% Kappa agreement (∼6.9-11.6% higher than other IAs) in detecting sleep epochs. The proposed deep learning model did not differ significantly from the reference EEG in estimating sleep onset latency (SOL), wake after sleep onset (WASO), total sleep time (TST), and sleep efficiency (SE). Amount of bias and minimum detectable change in estimating SOL, WASO, TST and SE by the deep learning model was smaller than other four IAs.
The proposed novel deep learning algorithm simultaneously incorporating ZCM/PIM mode data performs significantly better in assessing sleep than existing conventional IAs.
通过手腕活动记录仪估计睡眠参数高度依赖于将运动数据转换为睡眠/清醒评分的解释算法（IA）的性能。
（1）活动记录仪的操作模式——比例积分测量（PIM）或过零模式（ZCM），分别对应运动强度和频率——是否会影响睡眠评分；以及（2）能否通过结合PIM/ZCM数据的深度学习方法开发出性能更好的睡眠评分IA。
从一项单一的家庭夜间睡眠研究中获取了40名健康成年人（17名女性，平均年龄：26.7岁）的ZCM和PIM以及脑电图（EEG）数据。首先通过将几种经典深度学习模型应用于仅PIM、仅ZCM以及组合的ZCM/PIM数据来评估操作模式的影响。之后，开发了一种结合ZCM/PIM数据的新型深度学习模型，并将其性能与现有的科尔 - 克里普克、重新评分的科尔 - 克里普克、萨德和加州大学圣地亚哥分校的IA进行比较。
相对于EEG参考，ZCM/PIM组合模式在对睡眠/清醒时段进行评分时的一致性高于仅ZCM或PIM模式。所提出的新型深度学习模型在检测睡眠时段时显示出87.7%的准确率（比其他IA高0.2 - 1%）、94.1%的灵敏度（比其他IA低0.7 - 4.3%）、64.0%的特异性（比其他IA高9.9 - 21.5%）以及59.9%的卡帕一致性（比其他IA高约6.9 - 11.6%）。在估计睡眠开始潜伏期（SOL）、睡眠中觉醒（WASO）、总睡眠时间（TST）和睡眠效率（SE）方面，所提出的深度学习模型与参考EEG没有显著差异。深度学习模型在估计SOL、WASO、TST和SE时的偏差量和最小可检测变化量均小于其他四种IA。
所提出的同时结合ZCM/PIM模式数据的新型深度学习算法在评估睡眠方面的表现明显优于现有的传统IA。
