ausblenden:
Schlagwörter:
-
Zusammenfassung:
Recently, we proposed a framework for automatized, model-free MR sequence design for contrast generation. The concept was proven by optimizing the contrast-generating sequence using a real MRI scanner for scanning model solutions repeatedly until convergence to the target. In this work, we demonstrate for the first time that this live optimization loop can also be used directly on a real subject’s brain. As an example, GM/WM contrast for a GRE sequence was optimized fully automatically in a model-free and reference-free manner live in a healthy subject at a 3T MR scanner.
