Joint super-resolution and synthesis of 1 mm isotropic MP-RAGE volumes from clinical MRI exams with scans of different orientation, resolution and contrast
Juan Eugenio Iglesias,
Benjamin Billot,
Yaël Balbastre,
Azadeh Tabari,
John Conklin,
R. Gilberto González,
Daniel C. Alexander,
Polina Golland,
Brian L. Edlow,
Bruce Fischl
Affiliations
Juan Eugenio Iglesias
Corresponding author at: Centre for Medical Image Computing, Department of Medical Physics and Biomedical Engineering, University College London, London, UK.; Centre for Medical Image Computing, Department of Medical Physics and Biomedical Engineering, University College London, UK; Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital and Harvard Medical School, Boston, USA; Computer Science and Artificial Intelligence Laboratory, Massachusetts Institute of Technology, Boston, USA
Benjamin Billot
Centre for Medical Image Computing, Department of Medical Physics and Biomedical Engineering, University College London, UK
Yaël Balbastre
Centre for Medical Image Computing, Department of Medical Physics and Biomedical Engineering, University College London, UK
Azadeh Tabari
Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital and Harvard Medical School, Boston, USA; Department of Radiology, Massachusetts General Hospital, Boston, USA
John Conklin
Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital and Harvard Medical School, Boston, USA; Department of Radiology, Massachusetts General Hospital, Boston, USA
R. Gilberto González
Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital and Harvard Medical School, Boston, USA; Neuroradiology Division, Massachusetts General Hospital, Boston, USA
Daniel C. Alexander
Centre for Medical Image Computing, Department of Medical Physics and Biomedical Engineering, University College London, UK
Polina Golland
Computer Science and Artificial Intelligence Laboratory, Massachusetts Institute of Technology, Boston, USA
Brian L. Edlow
Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital and Harvard Medical School, Boston, USA; Center for Neurotechnology and Neurorecovery, Massachusetts General Hospital, Boston, USA
Bruce Fischl
Athinoula A. Martinos Center for Biomedical Imaging, Massachusetts General Hospital and Harvard Medical School, Boston, USA
Most existing algorithms for automatic 3D morphometry of human brain MRI scans are designed for data with near-isotropic voxels at approximately 1 mm resolution, and frequently have contrast constraints as well-typically requiring T1-weighted images (e.g., MP-RAGE scans). This limitation prevents the analysis of millions of MRI scans acquired with large inter-slice spacing in clinical settings every year. In turn, the inability to quantitatively analyze these scans hinders the adoption of quantitative neuro imaging in healthcare, and also precludes research studies that could attain huge sample sizes and hence greatly improve our understanding of the human brain. Recent advances in convolutional neural networks (CNNs) are producing outstanding results in super-resolution and contrast synthesis of MRI. However, these approaches are very sensitive to the specific combination of contrast, resolution and orientation of the input images, and thus do not generalize to diverse clinical acquisition protocols – even within sites. In this article, we present SynthSR, a method to train a CNN that receives one or more scans with spaced slices, acquired with different contrast, resolution and orientation, and produces an isotropic scan of canonical contrast (typically a 1 mm MP-RAGE). The presented method does not require any preprocessing, beyond rigid coregistration of the input scans. Crucially, SynthSR trains on synthetic input images generated from 3D segmentations, and can thus be used to train CNNs for any combination of contrasts, resolutions and orientations without high-resolution real images of the input contrasts. We test the images generated with SynthSR in an array of common downstream analyses, and show that they can be reliably used for subcortical segmentation and volumetry, image registration (e.g., for tensor-based morphometry), and, if some image quality requirements are met, even cortical thickness morphometry. The source code is publicly available at https://github.com/BBillot/SynthSR.
