Scholarlink

<p>The diagnosis of rare rheumatic diseases is challenging due to their complex and often atypical symptoms. Diagnostic delays are often associated with significant consequences for morbidity and increased costs. Traditional digital diagnostic decision support systems (DDSS) have only been able to close this gap to a&#xa0;limited extent. Large language models (LLMs), on the other hand, differ fundamentally in how they work, as they can process free text, enable simple questions to be asked, and process multimodal information. There is growing evidence that LLMs appear to be superior to conventional DDSS in various areas. This review classifies the current evidence, discusses the benefits and risks in rheumatological care, and outlines the prerequisites for safe implementation in clinical workflows, including regulatory and ethical aspects.</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

Digitale Diagnoseunterstützung seltener rheumatologischer Erkrankungen: Evidenz und Perspektiven

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年1月26日

查看全文
  • Phillip Kremer,
  • Martin Krusche,
  • Johannes Knitza

摘要

The diagnosis of rare rheumatic diseases is challenging due to their complex and often atypical symptoms. Diagnostic delays are often associated with significant consequences for morbidity and increased costs. Traditional digital diagnostic decision support systems (DDSS) have only been able to close this gap to a limited extent. Large language models (LLMs), on the other hand, differ fundamentally in how they work, as they can process free text, enable simple questions to be asked, and process multimodal information. There is growing evidence that LLMs appear to be superior to conventional DDSS in various areas. This review classifies the current evidence, discusses the benefits and risks in rheumatological care, and outlines the prerequisites for safe implementation in clinical workflows, including regulatory and ethical aspects.