Radiomics analysis enhances the diagnostic performance of CMR stress perfusion: a proof-of-concept study using the...

Area of Science:

• Cardiovascular Imaging
• Radiomics
• Artificial Intelligence in Medicine

Background:

• Coronary artery disease (CAD) diagnosis relies on invasive and non-invasive methods.
• Cardiovascular magnetic resonance (CMR) stress perfusion imaging is a key non-invasive tool.
• Assessing perfusion defects visually can be subjective and may miss subtle findings.

Purpose of the Study:

• To evaluate the feasibility of extracting radiomics features from CMR stress perfusion images.
• To compare the diagnostic performance of radiomics models against standard visual assessment.
• To determine if radiomics can improve the detection of myocardial perfusion defects.

Main Methods:

• Utilized the Dan-NICAD 1 dataset (148 patients) with CMR stress perfusion and invasive Fractional Flow Reserve (FFR).
• Extracted radiomics features (first-order and texture) from left ventricular myocardium using the Pyradiomics platform.
• Developed and validated radiomics models, comparing them to conventional qualitative visual analysis.

Main Results:

• Included 92 patients (39 with positive FFR).
• Conventional visual analysis showed 41% sensitivity and 84% specificity.
• The radiomics model achieved an average sensitivity of 53% and specificity of 86%.

Conclusions:

• Demonstrated the feasibility of radiomics analysis on CMR stress perfusion images.
• Radiomics models suggest superior diagnostic performance compared to visual assessment for detecting perfusion defects.
• This proof-of-concept study highlights radiomics as a promising tool for enhancing CAD diagnosis.