Artificial Intelligence for Contrast-Free MRI: Scar Assessment in Myocardial Infarction Using Deep Learning-Based...

Area of Science:

• Cardiovascular Imaging
• Artificial Intelligence in Medicine
• Medical Diagnostics

Background:

• Cardiovascular magnetic resonance late gadolinium enhancement (LGE) is the gold standard for assessing myocardial scars.
• A contrast-free approach for myocardial scar assessment offers significant advantages, including reduced scan time and cost, and elimination of contrast-associated risks.

Purpose of the Study:

• To introduce and evaluate Virtual Native Enhancement (VNE), a novel contrast-free technology for myocardial scar assessment.
• To compare the efficacy and image quality of VNE against the established LGE technique.

Main Methods:

• VNE utilizes artificial intelligence (generative adversarial networks) to create LGE-like images from cine imaging and native T1 maps.
• The VNE model was developed and validated using a large dataset of 3002 cardiac MRI scans from 775 patients with previous myocardial infarction.
• Scar quantification (volume and transmurality) was performed using established methods and compared between VNE and LGE.

Main Results:

• VNE demonstrated superior image quality compared to LGE in a blinded analysis by independent operators.
• VNE showed strong correlations with LGE in quantifying scar size (R=0.89) and transmurality (R=0.84).
• VNE achieved an 84% overall accuracy in scar detection compared to LGE, with 100% specificity and 77% sensitivity.

Conclusions:

• VNE is a highly accurate, contrast-free method for assessing myocardial scars, showing excellent agreement with LGE in terms of distribution and quantification.
• VNE offers superior image quality and has the potential to significantly reduce scan times and costs.
• This AI-based technology promises to enhance the accessibility and efficiency of cardiovascular magnetic resonance imaging for myocardial scar assessment.