München München

Digitalimpulse: Artificial intelligence agents for diagnostic purposes

ESC Congress 2026 | In this interview, Dr. Rohan Khera (Yale School of Medicine, New Haven, USA) discusses the future prospects of artificial intelligence agents for diagnostic purposes. What might be feasible applications for autonomous AI agents, and what role will clinicians play in their implementation and subsequent use?

By:

PD Dr. Philipp Breitbart

Section Head Digital Cardiology

 

 

2026-09-01

Image source (image above): Sergii Figurnyi / Shutterstock.com

Take-aways

  • Autonomous AI agents go beyond task-specific AI tools by combining information from different parts of the diagnostic process and activating multiple tools to address a clinical question.
  • Physicians should retain control over the final clinical decision—at least for now. AI could review complex patient information, propose decisions, and explain its reasoning, while the clinician remains responsible for the final step, such as ordering a test.
  • AI agents could help clinicians manage the complexity of patient records by bringing together data from labs, medications, ECGs, echocardiograms, and other sources to support a more comprehensive assessment.
  • The adoption of autonomous AI will require gradual implementation and clinician involvement. Tools may initially be deployed as simple applications, while centralized data infrastructure could enable remote management and monitoring; cardiologists should actively shape how future care evolves.

To the overview page ESC Congress 2026

Related news

Medizinische Lehre von morgen bzw. Lehre über digitale Medizin: Mit Prof. B. Meder und Dr. K. Betz.

ESC 2026 | TRIC-I-HF: Trikuspidalklappeninterventionen verbesserten klinische Endpunkte in einer Hochrisikopopulation. Kommentar von Dr. N. Wunderlich.

ESC 2026 | Prof. U. Bavendiek zu den Erkenntnissen der neuen Subanalyse der DIGIT-HF-Studie. Interview mit PD Dr. S. Perings.