Learning from a Handful Volumes: MRI Resolution Enhancement with Volumetric Super-Resolution Forests
Authors:
Preprint
Abstract
Magnetic resonance imaging (MRI) enables 3-D imaging of anatomical structures. However, the acquisition of MR volumes with high spatial resolution leads to long scan times. To this end, we propose volumetric super-resolution forests (VSRF) to enhance MRI resolution retrospectively. Our method learns a locally linear mapping between low-resolution and high-resolution volumetric image patches by employing random forest regression. We customize features suitable for volumetric MRI to train the random forest and propose a median tree ensemble for robust regression. VSRF outperforms state-of-the-art example-based super-resolution in term of image quality and efficiency for model training and inference in different MRI datasets. It is also superior to unsupervised methods with just a handful or even a single volume to assemble training data.Keywords
All Related Versions
This publication has 12 references indexed in Scilit:
- Deep Image Super Resolution via Natural Image Priors2018
- Brain MRI super resolution using 3D deep densely connected neural networks2018
- Bayesian Image Quality Transfer with CNNs: Exploring Uncertainty in dMRI Super-Resolution2017
- Robust Multiframe Super-Resolution Employing Iteratively Re-Weighted MinimizationIEEE Transactions on Computational Imaging, 2016
- Image Quality Transfer via Random Forest Regression: Applications in Diffusion MRI2014
- Single-image super-resolution of brain MR images using overcomplete dictionariesMedical Image Analysis, 2013
- Multi-parametric neuroimaging reproducibility: A 3-T resource studyNeuroImage, 2011
- Non-local MRI upsamplingMedical Image Analysis, 2010
- Image Super-Resolution Via Sparse RepresentationIEEE Transactions on Image Processing, 2010
- Image quality assessment: from error visibility to structural similarityIEEE Transactions on Image Processing, 2004