Hassan Bagher-Ebadian1,2, Ramesh Paudyal1,2, Tom Mikkelsen3, Quan Jiang1, James Russel Ewing1,2
1Neurology, Henry Ford Hospital, Detroit, MI, USA; 2Physics, Oakland University, Rochester, MI, USA; 3Neurosergery, Henry Ford Hospital, Detroit, MI, USA
In this study an adaptive neural network is employed for direct estimation of longitudinal relaxation time (T1) in Look-Locker signals. It is hypothesized that, given a signal generated by a LL model, an ANN could be trained to directly estimate T1. The analytical equation of the Look-Locker signal was considered as the gold standard of training and a set of LL signals for a wide range of T1s were generated. For each T1 value LL signal inputs were generated by varying the other independent parameters in the synthetic model of signal (T2*, M0 etc).