News

We are excited to share that two of our Computers Science Ph.D Students Bingyu Xin and Meng Ye were awarded two accolades by securing first place in both tasks of the Cardiac MRI Reconstruction (CMRxRecon) Challenge at Statistical Atlases and Computational Modeling of the Heart (STACOM) 2023 workshop. This workshop was held in conjunction with the MICCAI (best medical image analysis conference) 2023 in Vancouver, Canada.

In their work, presented in the paper with medical collaborators from New York University Grossman School of Medicine and titled "Fill the K-Space and Refine the Image: Prompting for Dynamic and Multi-Contrast MRI Reconstruction", they introduced a prompt learning-based all-in-one unrolled method for dynamic and multi-contrast/-view cardiac MRI reconstruction from highly undersampled multi-coil k-space raw data.  Their approach gives the ability to recover small anatomical structures for cardiac MRI better than the state-of-the-art, which has important significance for clinical diagnosis.

This method not only outperformed over 200 teams from 22 countries in the challenge but also achieved state-of-the-art performance on another well-known large-scale MRI reconstruction challenge dataset, the fastMRI for knee imaging.

The team was led by Distinguished Professor Dimitris Metaxas.