München München

Digitalimpulse: Rethinking Cardiovascular Risk

ESC Congress 2026 | In this interview, Prof. Harlan Krumholz (Yale School of Medicine, New Haven, USA) discusses how AI can improve individual cardiovascular risk prediction using multidimensional data and personalized risk assessment as well as crucial question of how AI can ultimately benefit patients.

By:

PD Dr. Philipp Breitbart

Section Head Digital Cardiology

 

 

2026-09-02

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

Take-aways

  • From broad risk scores to precision medicine: Traditional risk scores consider only a limited number of factors. AI and advanced analytics can integrate far more information about an individual patient.
  • Multidimensional data open new possibilities: Health records, behavior, exposures, biology and imaging can be combined to provide a more comprehensive picture of individual cardiovascular risk.
  • Clinical implementation is getting closer – but validation is key: AI-based models need to demonstrate that they work across different populations and clinical settings before they can be widely adopted.
  • AI should support, not replace, clinical judgment: When AI-based predictions conflict with a clinician’s assessment, the technology should be treated as a partner. Human judgment remains essential to incorporate clinical context and patients’ values and preferences.
  • Better prediction does not automatically mean better outcomes: The ultimate test of AI is not how accurately it predicts risk, but whether it leads to better decisions and, ultimately, better outcomes for patients.

To the overview page ESC Congress 2026

Related news

Wie kommt KI vom wissenschaftlichen Modell in die klinische Praxis? Mit Prof. C. Dlaska und PD Dr. P. Breitbart.

ESC 2026 | Prof. J. Bauersachs im Interview zum aktuellen Stand bei ASIs wie Baxdrostat, Lundostad und Vicatrostad.

In diesem Digital Wrap-up geht es um die digitalen Highlights vom ESC Congress in München, der sich dem Fokusthema Artificial Intelligence widmete.