Cesari Matteo, Stefani Ambra, Mitterling Thomas, Frauscher Birgit, Schönwald Suzana V, Högl Birgit
Department of Neurology, Medical University of Innsbruck, Innsbruck, Austria.
Department of Neurology, Medical University of Innsbruck, Innsbruck, Austria.
Sleep Med. 2021 Jan;77:136-146. doi: 10.1016/j.sleep.2020.11.033. Epub 2020 Dec 5.
In current clinical practice, sleep is manually scored in discrete stages of 30-s duration. We hypothesize that modelling sleep automatically as continuous and dynamic process predicts healthy ageing better than traditional scoring.
Sleep electroencephalography of 15 young healthy subjects (aged ≤40 years) was used to train the modelling method. Each 3-s sleep mini-epoch was modelled as a probabilistic combination of wakefulness, light and deep sleep. For 79 healthy sleepers (aged 20-77 years), 15 sleep features were derived from manual traditional scoring (manual features), 7 from the automatic modelling (automatic features) and 24 from a combination of automatic modelling with traditional scoring (combined features). Age was predicted with seven multiple linear regression models with i) manual, ii) automatic, iii) combined, iv) manual + automatic, v) manual + combined, vi) automatic + combined, and vii) manual + automatic + combined sleep features. Using the same seven sleep feature groups, two support vector machine and one random forest classifiers were used to discriminate younger (aged <47 years) from older subjects with fivefold cross-validation. Adjusted coefficients of determination (adj-R) and average validation accuracy (ACC) were used to compare the linear models and the classifiers.
The linear model and the classifiers using only manual features achieved the lowest values of adjusted coefficient of determination and classification validation accuracy (adj-R = 0.295, ACC = 63.00% ± 16.22%) compared to the ones using automatic (adj-R = 0.354, ACC = 65.83% ± 9.39%), combined (adj-R = 0.321, ACC = 63.42% ± 8.78%), manual + automatic (adj-R = 0.416, ACC = 67.00% ± 8.60%), manual + combined (adj-R = 0.355, ACC = 72.17% ± 12.90%), automatic + combined (adj-R = 0.448, ACC = 65.92% ± 7.97%), and manual + automatic + combined sleep features (adj-R = 0.464, ACC = 70.92% ± 10.33%).
Continuous and dynamic sleep modelling captures healthy ageing better than traditional sleep scoring.
在当前临床实践中，睡眠是按照30秒时长的离散阶段进行人工评分的。我们假设，将睡眠自动建模为一个连续且动态的过程，比传统评分方法能更好地预测健康衰老。
使用15名年轻健康受试者（年龄≤40岁）的睡眠脑电图来训练建模方法。每3秒的睡眠微时段被建模为清醒、浅睡眠和深睡眠的概率组合。对于79名健康睡眠者（年龄20 - 77岁），从传统人工评分中得出15个睡眠特征（人工特征），从自动建模中得出7个（自动特征），从自动建模与传统评分相结合的方式中得出24个（组合特征）。使用七个多元线性回归模型预测年龄，模型分别采用：i）人工、ii）自动、iii）组合、iv）人工 + 自动、v）人工 + 组合、vi）自动 + 组合、vii）人工 + 自动 + 组合睡眠特征。使用相同的七个睡眠特征组，通过五折交叉验证，采用两个支持向量机和一个随机森林分类器来区分年轻受试者（年龄<47岁）和年长受试者。使用调整后的决定系数（adj-R）和平均验证准确率（ACC）来比较线性模型和分类器。
与使用自动特征（adj-R = 0.354，ACC = 65.83% ± 9.39%）、组合特征（adj-R = 0.321，ACC = 63.42% ± 8.78%）、人工 + 自动特征（adj-R = 0.416，ACC = 67.00% ± 8.60%）、人工 + 组合特征（adj-R = 0.355，ACC = 72.17% ± 12.90%）、自动 + 组合特征（adj-R = 0.448，ACC = 65.92% ± 7.97%）以及人工 + 自动 + 组合睡眠特征（adj-R = 0.464，ACC = 70.92% ± 10.33%）的模型相比，仅使用人工特征的线性模型和分类器的调整决定系数和分类验证准确率最低（adj-R = 0.295，ACC = 63.00% ± 16.22%）。
连续且动态的睡眠建模比传统睡眠评分能更好地捕捉健康衰老情况。
