BeatMapSeq: A Framework for In-Line Beat-Based Cardiac Parametric Mapping Sequence Construction and Map Building.

Area of Science:

• Cardiovascular Magnetic Resonance Imaging
• Medical Physics
• Biomedical Engineering

Background:

• Cardiac parametric mapping provides crucial physiological information but often involves complex sequence development and map generation.
• Existing methods for creating cardiac mapping sequences and building maps can be time-consuming and require specialized programming skills.

Purpose of the Study:

• To introduce BeatMapSeq, a novel framework for simplifying sequence construction and map building in cardiac parametric mapping using cardiovascular magnetic resonance.
• To validate the efficiency and accuracy of BeatMapSeq's components, BeatMap Framework and DeepFittingNet, for various cardiac mapping techniques.

Main Methods:

• Developed BeatMap Framework for interactive, no-programming cardiac parametric mapping sequence construction on the scanner.
• Integrated DeepFittingNet, a neural network for automated, in-line generation of parametric maps.
• Validated BeatMapSeq on Philips 3.0-T scanners using in vivo studies for T1, T2, T1rho, and simultaneous T1/T2 mapping sequences.

Main Results:

• All tested cardiac parametric mapping sequences were successfully constructed interactively on the scanner without programming.
• BeatMapSeq's MOLLI implementation showed excellent agreement with vendor-provided sequences.
• DeepFittingNet generated maps automatically with good agreement to nonlinear least-squares (NLS) fitting, enabling rapid availability of parametric maps.

Conclusions:

• BeatMapSeq significantly streamlines the process of cardiac parametric mapping in cardiovascular magnetic resonance.
• The framework facilitates efficient sequence design and automated map generation, making advanced cardiac imaging more accessible.
• BeatMapSeq offers a robust solution for in-line map building, improving workflow efficiency for cardiovascular magnetic resonance examinations.