Model-informed machine learning for multi-component T2 relaxometry

Thomas Yu, Erick Jorge Canales-Rodríguez, Marco Pizzolato, Gian Franco Piredda, Tom Hilbert, Elda Fischi Gomez, Matthias Weigel, Muhamed Barakovic, Meritxell Bach Cuadra, Cristina Granziera, Tobias Kober, Jean-Philippe Thiran

Abstract: Highlights•Recovering T2<math><msub is="true"><mi is="true">T</mi><mn is="true">2</mn></msub></math> distributions from MRI data is difficult, especially due to noise.•Noise robust recovery possible through machine learning on synthetic data.•Synthetic data is derived from biophysical models.•Validated on in-vivo/ex-vivo scans including a subject with multiple sclerosis.