Nonlinear Equivariant Imaging: Learning Multi-Parametric Tissue Mapping without Ground Truth for Compressive Quantitative MRI
This addresses the limitation of needing high-fidelity ground truth for deep MRF reconstruction in medical imaging, offering a more practical solution for quantitative MRI.
The paper tackles the problem of reconstructing quantitative tissue maps from compressive Magnetic Resonance Fingerprinting (MRF) without requiring ground truth data, by proposing NonLinear Equivariant Imaging (NLEI), a self-supervised learning approach that closely approaches supervised learning performance in tests on two acquisition settings.
Current state-of-the-art reconstruction for quantitative tissue maps from fast, compressive, Magnetic Resonance Fingerprinting (MRF), use supervised deep learning, with the drawback of requiring high-fidelity ground truth tissue map training data which is limited. This paper proposes NonLinear Equivariant Imaging (NLEI), a self-supervised learning approach to eliminate the need for ground truth for deep MRF image reconstruction. NLEI extends the recent Equivariant Imaging framework to nonlinear inverse problems such as MRF. Only fast, compressed-sampled MRF scans are used for training. NLEI learns tissue mapping using spatiotemporal priors: spatial priors are obtained from the invariance of MRF data to a group of geometric image transformations, while temporal priors are obtained from a nonlinear Bloch response model approximated by a pre-trained neural network. Tested retrospectively on two acquisition settings, we observe that NLEI (self-supervised learning) closely approaches the performance of supervised learning, despite not using ground truth during training.