Skrøvseth Stein Olav, Arsand Eirik, Godtliebsen Fred, Joakimsen Ragnar M
Norwegian Centre for Integrated Care and Telemedicine, University Hospital of North Norway, Tromsø, Norway.
Stud Health Technol Inform. 2012;180:1045-9.
We gathered a data set from 30 patients with type 1 diabetes by giving the patients a mobile phone application, where they recorded blood glucose measurements, insulin injections, meals, and physical activity. Using these data as a learning data set, we describe a new approach of building a mobile feedback system for these patients based on periodicities, pattern recognition, and scale-space trends. Most patients have important patterns for periodicities and trends, though better resolution of input variables is needed to provide useful feedback using pattern recognition.
我们通过为30名1型糖尿病患者提供一款手机应用程序来收集数据集，患者可在该程序中记录血糖测量值、胰岛素注射情况、饮食和身体活动。将这些数据用作学习数据集，我们描述了一种基于周期性、模式识别和尺度空间趋势为这些患者构建移动反馈系统的新方法。大多数患者具有周期性和趋势的重要模式，不过需要更好地解析输入变量，以便使用模式识别提供有用的反馈。
