Scholarlink

<p>Artificial intelligence (AI) is gaining importance in the field of cardiology. By analyzing complex multimodal data AI can support the diagnostic processes, risk stratification and making decisions. In cardiac imaging AI-based procedures improve the objective image analysis and facilitate the detection of specific cardiac diseases. The AI-assisted analysis in electrocardiography (ECG) not only enables reliable identification of established diagnoses but also enables the detection of patterns not yet identifiable for the human eye and therefore opens up opportunities for the establishment of new diagnostics and risk stratification. Large language models (LLM) represent a&#xa0;subdomain of AI. In cardiology they show particular potential for knowledge-based and text-based tasks, such as clinical documentation, support in drafting medical reports and patient education. Based on the current evidence, publicly available LLM are not appropriate for image interpretation, clinical decision-making or for supporting peer-review processes. Key challenges for the clinical implementation of AI include data protection and data security, regulatory requirements, currently limited evidence from randomized clinical trials, the limited option for validation of self-learning algorithms, limited transparency and the need for structured training programs for both physicians and patients. Irrespective of the degree of AI assistance, physician oversight remains crucial.</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

Künstliche Intelligenz in der Inneren Medizin am Beispiel der Kardiologie

  • 期刊论文

  • 发表日期：

    2026年4月28日

查看全文
  • Henrike A. K. Hillmann,
  • Amalia Sluha,
  • David Duncker

摘要

Artificial intelligence (AI) is gaining importance in the field of cardiology. By analyzing complex multimodal data AI can support the diagnostic processes, risk stratification and making decisions. In cardiac imaging AI-based procedures improve the objective image analysis and facilitate the detection of specific cardiac diseases. The AI-assisted analysis in electrocardiography (ECG) not only enables reliable identification of established diagnoses but also enables the detection of patterns not yet identifiable for the human eye and therefore opens up opportunities for the establishment of new diagnostics and risk stratification. Large language models (LLM) represent a subdomain of AI. In cardiology they show particular potential for knowledge-based and text-based tasks, such as clinical documentation, support in drafting medical reports and patient education. Based on the current evidence, publicly available LLM are not appropriate for image interpretation, clinical decision-making or for supporting peer-review processes. Key challenges for the clinical implementation of AI include data protection and data security, regulatory requirements, currently limited evidence from randomized clinical trials, the limited option for validation of self-learning algorithms, limited transparency and the need for structured training programs for both physicians and patients. Irrespective of the degree of AI assistance, physician oversight remains crucial.