Magnetic Resonance Fingerprinting Using Recurrent Neural Networks

摘要: Magnetic Resonance Fingerprinting (MRF) is a new approach to quantitative magnetic resonance imaging that allows simultaneous measurement of multiple tissue properties in single, time-efficient acquisition. Standard MRF reconstructs parametric maps using dictionary matching and requires high computational time. We propose perform map reconstruction recurrent neural network, which exploits the time-dependent information signal evolution. evaluate our method on multiparametric synthetic signals compare it existing approaches, including those based networks. Our achieves state-of-the-art estimates T1 T2 values. In addition, time reduced compared dictionary-matching approach.