Bianchi Anna M, Mendez Martin O
Politecnico di Milano, IT 20133 Italia.
Annu Int Conf IEEE Eng Med Biol Soc. 2010;2010:3606-9. doi: 10.1109/IEMBS.2010.5627432.
In the present work we apply a fully automatic procedure to the analysis of signal coming from a sensorized T-shit, worn during the night, for sleep evaluation. The goodness and reliability of the signals recorded trough the T-shirt was previously tested, while the employed algorithms for feature extraction and sleep classification were previously developed on standard ECG recordings and the obtained classification was compared to the standard clinical practice based on polysomnography (PSG). In the present work we combined T-shirt recordings and automatic classification and could obtain reliable sleep profiles, i.e. the sleep classification in WAKE, REM (rapid eye movement) and NREM stages, based on heart rate variability (HRV), respiration and movement signals.
在本研究中，我们应用一种全自动程序来分析夜间穿着的装有传感器的T恤所传来的信号，以进行睡眠评估。此前已测试了通过T恤记录的信号的质量和可靠性，而用于特征提取和睡眠分类的算法此前是基于标准心电图记录开发的，并且将所获得的分类结果与基于多导睡眠图（PSG）的标准临床实践进行了比较。在本研究中，我们将T恤记录与自动分类相结合，基于心率变异性（HRV）、呼吸和运动信号，能够获得可靠的睡眠概况，即清醒、快速眼动（REM）和非快速眼动（NREM）阶段的睡眠分类。
