Multi-stage sleep classification using photoplethysmographic sensor

The conventional approach to monitoring sleep stages requires placing multiple sensors on the patients, which is inconvenient for long-term monitoring and requires expert support. We propose a single sensor Photoplethysmographic (PPG) based automated multi-stage sleep classification. This experimental study recorded the PPG during the entire night's sleep of ten patients. Data analysis was performed to obtain 82 features from the recordings, which were then classified against the sleep stages. The classification results using SVM with the polynomial kernel gave the overall accuracy of 84.66%, 79.62%, and 72.23% for two, three, and four-stage sleep classification. These results show that using only PPG; it is possible to conduct sleep stage monitoring. These findings open the opportunities for PPG-based wearable solutions for home-based automated sleep monitoring., The PSG data were recorded for the night sleep duration of ten participants (9 male/ 1 female, age 43–75 years). The length of sleep time ranged from 6.8 to 10.1 hours. All participants were volunteers and recruited from the out-patients at Charite Hospital, Berlin, Germany. All suffered sleep-disordered breathing and were free from a history of cardiac issues. The diagnosis was based on PSG outcomes and clinical symptoms. The research and data collection protocol was approved by the Charite Hospital Committee for Ethics in Human Research (2018), Berlin, Germany, and the experiments were conducted in accordance with the Helsinki declaration for ethical experiments, revised in 2013. Written consent was taken prior to the experiments. The demographic information of the subjects. Each PSG recording included two-channel EEG (channel C3-A2 and C4-A1), ECG, PPG, left and right EOG, leg movements, thoracic and abdominal wall expansion, arterial oxygen saturation SaO2, and oronasal airflow. Acc..., Microsoft Excel or Matlab or Python.

传统睡眠分期监测方法需在患者体表布设多组传感器，不仅不利于长期监测，还需要专业人员辅助操作。本研究提出一种基于单传感器光电容积描记法（Photoplethysmographic, PPG）的自动化多阶段睡眠分类方案。本实验记录了10名患者整夜睡眠期间的PPG信号，通过对采集数据进行分析提取得到82项特征，并以此开展睡眠分期分类任务。采用搭载多项式核的支持向量机（Support Vector Machine, SVM）进行分类，在两阶段、三阶段与四阶段睡眠分类任务中，整体准确率分别达到84.66%、79.62%与72.23%。研究结果表明，仅依靠PPG信号即可实现睡眠分期监测，这为基于PPG的可穿戴设备应用于居家自动化睡眠监测提供了新的可行方向。
本研究为10名参与者（9名男性、1名女性，年龄43~75岁）记录了夜间睡眠期间的多导睡眠图（Polysomnography, PSG）数据。参与者的总睡眠时间介于6.8至10.1小时之间。所有参与者均为志愿者，招募自德国柏林夏里特医院（Charite Hospital）的门诊患者，均患有睡眠呼吸障碍且无心脏疾病史。诊断依据PSG结果与临床症状共同确定。本研究的研究方案与数据采集流程已获得德国柏林夏里特医院人类研究伦理委员会批准（编号：2018），实验严格遵循2013年修订版《赫尔辛基宣言》的伦理实验规范。实验开展前已获取所有参与者的书面知情同意书。以下为受试者的人口统计学信息。每段PSG记录包含双通道脑电图（Electroencephalogram, EEG，通道C3-A2与C4-A1）、心电图（Electrocardiogram, ECG）、PPG、左右眼电图（Electrooculogram, EOG）、腿部运动监测、胸腹壁扩张度监测、动脉血氧饱和度（SaO2）以及口鼻气流监测。相关数据分析可通过Microsoft Excel、Matlab或Python完成。