Scholarlink

Background <p>Against the backdrop of increasing patient volumes, rising case complexity, and physicians′ limited time, AI-driven systems for anamnesis, triage, and documentation offer substantial potential for efficiency gains in medical history taking and clinical communication. Their development spans from rule-based decision trees to machine learning and large language model (LLM) dialogues, and further to “ambient” documentation that records the physician–patient interaction, autonomously extracts relevant information, and generates structured notes. Applications can be categorized by level of interaction (patient-, physician-, or system-facing) and by stage of care (before, during, and after the visit).</p> Practice <p>In routine practice, there is an evident shift away from generic symptom checkers—focused on safety but limited in diagnostic accuracy—toward domain-specific, curated intake and triage tools that demonstrate higher process relevance in elective care. The most immediate benefits currently arise from ambient documentation assistants: reduced typing workload, shorter post-visit processing times, and more complete, structured notes—always subject to final physician review and responsibility. Responsible deployment requires adherence to regulatory frameworks and compliance with MDR and the EU AI Act. A&#xa0;locally piloted example is “OrthoCopilot,” an adaptive, offline intake system generating structured summaries for physicians′ preparation.</p> Requirements <p>Stepwise introduction of AI-based technologies should be guided by measurable performance indicators (consultation time, workload, completeness, correction effort, billing quality), comply with prevailing regulations, enable early efficiency gains, support institutions in building internal expertise, and prepare the ground for the transition toward more advanced multimodal AI models.</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

Einsatz von künstlicher Intelligenz zur Next-Gen-Anamnese und Kommunikation in der O & U

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年5月8日

查看全文
  • Marco-Christopher Rupp,
  • Alexandros Doucas,
  • Sebastian Siebenlist

摘要

Background

Against the backdrop of increasing patient volumes, rising case complexity, and physicians′ limited time, AI-driven systems for anamnesis, triage, and documentation offer substantial potential for efficiency gains in medical history taking and clinical communication. Their development spans from rule-based decision trees to machine learning and large language model (LLM) dialogues, and further to “ambient” documentation that records the physician–patient interaction, autonomously extracts relevant information, and generates structured notes. Applications can be categorized by level of interaction (patient-, physician-, or system-facing) and by stage of care (before, during, and after the visit).

Practice

In routine practice, there is an evident shift away from generic symptom checkers—focused on safety but limited in diagnostic accuracy—toward domain-specific, curated intake and triage tools that demonstrate higher process relevance in elective care. The most immediate benefits currently arise from ambient documentation assistants: reduced typing workload, shorter post-visit processing times, and more complete, structured notes—always subject to final physician review and responsibility. Responsible deployment requires adherence to regulatory frameworks and compliance with MDR and the EU AI Act. A locally piloted example is “OrthoCopilot,” an adaptive, offline intake system generating structured summaries for physicians′ preparation.

Requirements

Stepwise introduction of AI-based technologies should be guided by measurable performance indicators (consultation time, workload, completeness, correction effort, billing quality), comply with prevailing regulations, enable early efficiency gains, support institutions in building internal expertise, and prepare the ground for the transition toward more advanced multimodal AI models.