Artificial Intelligence in CMR – Powerful Tools Require Careful Validation

Artificial intelligence is increasingly supporting everyday clinical CMR, particularly by reducing the time required for image analysis and post-processing.

However, different AI algorithms do not necessarily produce identical quantitative results. Understanding these differences is essential before AI-derived measurements are incorporated into clinical decision-making.

This important topic is addressed in a recent publication by Thomas Hadler et al., originating from the Berlin-Buch CMR research group, with several Circle Institute faculty members contributing as co-authors.

The study underlines the importance of quality assurance and independent validation as AI becomes part of routine clinical practice.