Henriques Teresa, Munshi Medha N, Segal Alissa R, Costa Madalena D, Goldberger Ary L
Wyss Institute for Biologically Inspired Engineering, Harvard University, Boston, MA, USA Center for Research in Health Technologies and Information Systems, Porto, Portugal Instituto de Telecomunicações, Porto, Portugal.
Beth Israel Deaconess Medical Center, Harvard Medical School, Boston, MA, USA Joslin Diabetes Center, Boston, MA, USA Harvard Medical School, Boston, MA, USA.
J Diabetes Sci Technol. 2014 Mar;8(2):299-306. doi: 10.1177/1932296814524095. Epub 2014 Mar 2.
The standard continuous glucose monitoring (CGM) output provides multiple graphical and numerical summaries. A useful adjunct would be a visualization tool that facilitates immediate assessment of both long- and short-term variability. We developed an algorithm based on the mathematical method of delay maps to display CGM signals in which the glucose value at time t is plotted against its value at time t. The data points are then color-coded based on their frequency of occurrence (density). Examples of this new visualization tool, along with the accompanying time series, are presented for selected patients with type 2 diabetes and non-diabetic controls over the age of 70 years. The method reveals differences in the structure of the glucose variability between subjects with a similar range of glucose values. We also observe that patients with comparable hemoglobin A1c (HbA1c) values may have very different delay maps, consistent with marked differences in the dynamics of glucose control. These differences are not accounted by the amplitude of the fluctuations. Furthermore, the delay maps allow for rapid recognition of hypo- and hyperglycemic periods over the full duration of monitoring or any subinterval. The glucose-at-a-glance visualization tool, based on colorized delay maps, provides a way to quickly assess the complex data acquired by CGM systems. This method yields dynamical information not contained in single summary statistics, such as HbA1c values, and may also serve as the basis for developing novel metrics of glycemic control.
标准的连续血糖监测（CGM）输出提供了多种图形和数值总结。一个有用的辅助工具将是一种可视化工具，它有助于对长期和短期变异性进行即时评估。我们基于延迟映射的数学方法开发了一种算法，用于显示CGM信号，其中将时间t的血糖值与其在时间t的血糖值进行绘制。然后根据数据点的出现频率（密度）对其进行颜色编码。针对选定的70岁以上2型糖尿病患者和非糖尿病对照，展示了这种新可视化工具的示例以及相应的时间序列。该方法揭示了血糖值范围相似的受试者之间血糖变异性结构的差异。我们还观察到，糖化血红蛋白（HbA1c）值相当的患者可能具有非常不同的延迟映射，这与血糖控制动态中的显著差异一致。这些差异不能通过波动幅度来解释。此外，延迟映射允许在整个监测期间或任何子区间内快速识别低血糖和高血糖期。基于彩色延迟映射的“一目了然”血糖可视化工具提供了一种快速评估CGM系统获取的复杂数据的方法。这种方法产生了单一汇总统计数据（如HbA1c值）中不包含的动态信息，并且还可能作为开发新型血糖控制指标的基础。
