Satellite system for observation of infants with inverted day-night sleeping cycles and their primary caregivers

为了观察婴儿的睡眠-清醒状态并识别这些状态的特征，人们致力于开发能够客观收集数据并进行分析的软件。本系统以婴儿身体运动为研究对象，采用视频睡眠描记术进行观察和分析。2004年，系统开发完成，软件基本完成。延时录像装置/系统改为数码相机和电脑，减轻了重量，并能够获取数字数据。2005年，为了验证软件的可靠性，进行了一次试验，软件读取了已经分析过的视频数据，并将早期的结果与新软件系统的分析结果进行了比较。然而，模拟视频数据的图像质量粗糙，难以进行精细分析，因此停止了分析。因此，视频被数字化仅用于遗留迁移。为了保存图像数据，使用一个硬盘连接到所需的大容量局域网。2006年，为了验证软件的可靠性，在收集婴儿数据时也使用了Actiwatch。对9例产后1个月婴儿的数据进行分析，在平均记录时间9.3小时中，除婴儿直接运动外，运动检测的激活率平均为6.0%。通过Spearman等级相关系数证明了Actiwatch的可靠性。在整个系统开发期间，笔者参加了国内外睡眠研究会议等，并通过演讲获得了睡眠研究者的反馈。在与海外联合研究员Thomas F. Anders博士（加州大学戴维斯分校精神病学和行为科学系，M.I.N.D.研究所）进行的试验中，该系统得到了改进。进一步明确的开发问题是自动分析方法，以及语音数据和重要数据（呼吸）的添加。

In order to observe the sleep-awake states of infants and identify the characteristics of these states, efforts were devoted to developing software that can objectively collect data and analyze it. This system focused on infant body movement, based on observation and analysis methods using videosomnography.In 2004, the system was developed and the software almost completed. The time-lapse video set/system was changed to a digital camera and computer, reducing weight and enabling acquisition of digital data.In 2005, a trial was conducted to validate the reliability of the software, where already-analyzed video data was read by the software and the earlier results were compared against the analysis results of the new software system. However, the image quality of the analogue video data was rough and subtle analysis was difficult, so analysis was stopped. Accordingly, the video was digitalized for legacy migration only. To save the image data, a hard disk was used connected to the required large capacity LAN.In 2006, in order to validate the reliability of the software, an Actiwatch was also used while collecting infant data. When data of 9 infants one month postpartum was analyzed, motion detection was activated other than for direct infant movement for an average 6.0% out of an average recording time of 9.3 hours. Reliability using the Actiwatch was demonstrated by Spearman's rank correlation coefficient.During the whole development period of this system, the author participated in domestic and overseas sleep research conferences, etc. and through presentations received feedback from sleep researchers. In trials conducted with overseas joint researcher Dr. Thomas F. Anders (Department of Psychiatry and Behavioral Sciences, UC Davis, M.I.N.D. Institute), the system was refined. Further development issues that became clear are the method of automatic analysis, and the addition of voice data and vital data (breathing).