RNN architecture with quantile layer reduces T1 and T2 reconstruction errors by over 80% versus prior CNNs on in-vivo brain MR fingerprinting data from multiple volunteers.
Nature Machine Intelligence
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
eess.IV 1years
2019 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
RinQ Fingerprinting: Recurrence-informed Quantile Networks for Magnetic Resonance Fingerprinting
RNN architecture with quantile layer reduces T1 and T2 reconstruction errors by over 80% versus prior CNNs on in-vivo brain MR fingerprinting data from multiple volunteers.