Digital Diabetology: Using a Digital Twin to Optimize Diabetes Management

根据国际糖尿病联合会(IDF)的数据，全世界有超过4.6亿人受到糖尿病的影响，相当于人口的近10%。数字化糖尿病管理可以为改善这些患者的治疗和生活质量做出重要贡献。尽管目前已经有了连续血糖监测(CGM)、智能手机或胰岛素泵等技术，但必须为个性化数字治疗支持开发新的概念。为此，该项目正专注于开发一对糖尿病双胞胎。它的核心是由代谢动力学的电子模型形成的，该模型还包括CGM传感器动力学和患者的食物摄入和运动。申请人已经建立了这些模型，并在一些研究中使用这些模型来分析活体动物模型的观察结果。这些技能将逐步添加到数字双胞胎中：包括对当前代谢状态的实时估计，以及对葡萄糖和胰岛素轨迹的模拟和预测。在回溯测量的基础上，也有可能使数字双胞胎适应患者的个体特征。这将提高数字孪生兄弟的质量，从而也可以得出诊断结论。该项目的研究目标是调查数字双胞胎的概念是否可以转移到糖尿病学，并在那里以一种有利于优化糖尿病管理的方式使用。为此目的，将首先在当前文献的背景下实施和评估原型。在对动物模型进行回顾分析的背景下，这一概念可以基于受控代谢测试，如口服葡萄糖耐量测试(OGTT)来验证。随后，以各种形式的糖尿病为例，对数字双胞胎进行了临床推广。因此，通过临床相关参数，如空腹血糖、HOMA、HbA1c、体重、体重指数(BMI)、总胆固醇、低密度脂蛋白(LDL)等对孪生子进行扩展，并对其预测质量进行分析。最后，提出了一个基于数字双胞胎的数字治疗支持系统的总体设计方案。这个项目是由梅德教授主持的。克劳迪娅·埃伯利来自富尔达应用科学大学。

According to the International Diabetes Federation (IDF), more than 460 million people worldwide are affected by diabetes mellitus, corresponding to a prevalence of almost 10% of the population. Digital diabetes management could make an important contribution to improve the therapy and quality of life of these patients. Although technologies such as continuous glucose monitoring (CGM), smartphones or insulin pumps are available today, new concepts must be developed for individual digital therapy support.To this end, the project is focusing on the development of a diabetological diabetic twin. Its core is formed by in silico models of metabolic dynamics that also include CGM sensor dynamics and the patient's food intake and movement. These models have already been established and used by the applicant in a number of studies to analyze observations from in vivo animal models. Skills will be progressively added to the digital twin: these include real-time estimation of current metabolic state and simulation and prediction of glucose and insulin trajectories. Based on retrospective measurements, it will also be possible to adapt the digital twin to the individual characteristics of the patient. This will improve the quality of the digital twin, which will then also allow diagnostic conclusions. The research goal of the project is to investigate if the concept of a digital twin can be transferred to diabetology and used there in a beneficial way for optimized diabetes management. For this purpose, a prototype will first be put into operation and assessed in the context of the current literature. In the context of a retrospective analysis of animal models, the concept can be validated on the basis of controlled metabolic tests such as the Oral Glucose Tolerance Test (OGTT). This is followed by the clinical extension of the digital twin using the example of various forms of diabetes mellitus. Therefore, the twin is extended by clinically relevant parameters, e.g. fasting glucose, HOMA, HbA1c, body weight, body mass index (BMI), total cholesterol, low density lipoprotein (LDL), etc., and its predictive quality is analyzed. Finally, an overall concept for the design of a digital therapy support based on a digital twin is proposed. The project is conducted by Prof. Dr. med. Claudia Eberle at Fulda University of Applied Sciences.