Reducing Phase Encoding Steps in Undersampled MRI k-Space using Novel Deep Learning Algorithms
By Sauman Das
Acquiring MRI scans is a time-taking and uncomfortable process. Any patient movement during the acquisition time can result in low quality scans which cause inaccurate diagnoses and problems during surgical treatment. In this study, I plan to study different signal undersampling procedures for k-space acquisition and develop a generative deep learning model to interpolate missing details. If this algorithm works successfully, the MRI acquisition process can be made more efficient and help millions of patients across the world.
Presentation Video (FCPS Only)