Robust Cardiac T1 Mapping in Three Heartbeats With UltraMAP: A Synergistic CNN-FC Approach

This paper proposes an ultrafast and robust myocardial T1 mapping technique for myocardial T1 mapping utilizing a minimal set of T1-weighted images obtained from a single inversion recovery (IR) pulse. By integrating a Convolutional Neural Network (CNN) with a Fully Connected (FC) network, UltraMAP proficiently handles three undersampled T1-weighted images acquired at distinct inversion-recovery times. This integration enables robust and precise myocardial T1 mapping, demonstrating remarkable noise immunity and reducing the need for extensive post-processing, thus advancing cardiac MRI diagnostics.