Scholarlink

<p>Artificial intelligence (AI) has the potential to fundamentally change diabetes care and substantially advance developments in the areas of prevention, prognostics, diagnostics, therapy, and treatment towards precision medicine. For example, AI can use voice characteristics, genetics, and imaging data for individual diabetes risk stratification in a&#xa0;preventive manner. By combining multimodal data and fundus images, AI enables early diagnosis of diabetes-related complications such as retino-, nephro- or neuropathy. AI-based systems are already able to support diabetes therapy through personalized nutritional analysis, automated carbohydrate estimation, medication selection, and insulin dosing. Especially in combination with continuous glucose monitoring (CGM) or automated insulin delivery (AID) systems, AI-based decision support solutions have demonstrated their clinical benefits by improving glycemic control and promoting adherence. In addition, AI is able to predictively detect hypoglycemia with the help of wearables or even in the vehicle cockpit and, thus, prevent emergency situations or traffic accidents. In the future, AI will also open up new avenues in drug development, biomarker identification and simulation-based decision-making with the help of digital twins. Current studies and application examples clearly point out that AI will not just be a&#xa0;simple add-on, but will develop towards a&#xa0;central tool for precise, personalized and proactive diabetes care.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Das Potenzial von künstlicher Intelligenz in der Diabetologie

  • 期刊论文

  • 发表日期：

    2025年8月19日

查看全文
  • Susanne Tan,
  • Young Hee Lee-Barkey,
  • Bernhard Kulzer,
  • Norbert Hermanns,
  • Dominic Ehrmann

摘要

Artificial intelligence (AI) has the potential to fundamentally change diabetes care and substantially advance developments in the areas of prevention, prognostics, diagnostics, therapy, and treatment towards precision medicine. For example, AI can use voice characteristics, genetics, and imaging data for individual diabetes risk stratification in a preventive manner. By combining multimodal data and fundus images, AI enables early diagnosis of diabetes-related complications such as retino-, nephro- or neuropathy. AI-based systems are already able to support diabetes therapy through personalized nutritional analysis, automated carbohydrate estimation, medication selection, and insulin dosing. Especially in combination with continuous glucose monitoring (CGM) or automated insulin delivery (AID) systems, AI-based decision support solutions have demonstrated their clinical benefits by improving glycemic control and promoting adherence. In addition, AI is able to predictively detect hypoglycemia with the help of wearables or even in the vehicle cockpit and, thus, prevent emergency situations or traffic accidents. In the future, AI will also open up new avenues in drug development, biomarker identification and simulation-based decision-making with the help of digital twins. Current studies and application examples clearly point out that AI will not just be a simple add-on, but will develop towards a central tool for precise, personalized and proactive diabetes care.