Open Resources for Advanced Imaging Research
CAI2R provides MRI data, tools for clinical translation, image reconstruction code, RF simulation tools, and image analysis software.
We also offer RF hardware.
To enable and encourage innovation throughout the field of biomedical imaging, we openly share advanced research resources. Many resources can be downloaded through this website; Some are hosted on dedicated micro-sites or maintained on version-controlled repositories, such as Github.
Jump to:
- Data
- Tools for clinical translation
- Hardware
- Reconstruction code
- Radiofrequency and dielectric tools
- Image analysis software
- Other tools
Data
Our fastMRI dataset—the world’s largest collection of its kind—has set the standard for the sharing of voluminous, curated, deidentified data for machine learning research on image reconstruction. Recently augmented with prostate MRI data, fastMRI is a dynamic resource that continues to grow, and has the potential to become the imaging research community’s premier hub for data interchange.
We are working on further additions, including GRASP data, longitudinal MRI data, and mesoscopic MRI data.
Raw k-space data and DICOM images from thousands of MRI scans of the knee, brain, and prostate, curated for machine learning research on image reconstruction.
Reproducibility data for two seminal machine learning MR image reconstruction studies.
Simultaneously acquired positron emission tomography and magnetic resonance data of the brain, plus joint image reconstruction code.
MRI data from suspensions of spherical polystyrene microbeads for research on gradient-echo and spin-echo signal decay in the presence of microstructural sources of magnetic susceptibility.
Tools for Clinical Translation
We provide free, open-source tools to support clinical-translation imaging studies. Mercure enables DICOM routing, automated image analysis, and AI integration. Yarra simplifies collection and management of raw data, workflow analysis, and image reconstruction.
In order to cover the full lifecycle of data acquisition, reconstruction, processing, and visualization, we are also developing a pulse sequence development tool for education and prototyping called mtrk, and a viewer for visualization of dynamic and high-dimensional datasets called Gravis.
A platform for integrating algorithms, AI models, and post-processing tools into clinical practice.
A platform for clinical translation of novel MRI reconstructions, raw data management, and imaging workflow analysis.
Hardware
A compact, stand-alone device for tracking and correcting patient motion in MRI exams.
An electronic interface for research and development on radio frequency coils built for Siemens Healthineers Tim 4G.
Reconstruction Code
An environment for GPU-accelerated linear and non-linear image reconstruction.
Simultaneous machine learning optimization of parallel MRI sampling pattern and variational-network image reconstruction parameters.
Machine learning optimization of k-space sampling for accelerated MRI.
State-of-the-art denoising for diffusion and functional imaging applications.
Software for robust estimation of DKI metrics
Uniform and robust fat/water separation for free-breathing radial MRI.
Software for T2 quantification in multi-spin-echo MRI.
-
Up to 40x acceleration for multi-coil gridding of arbitrary-shaped k-space data.
-
Reconstruction code for fast and flexible free-breathing dynamic volumetric MRI.
-
A high-quality reconstruction of accelerated MRI via joint estimation of image and coil sensitivities.
-
A nonlinear-optimization-based image reconstruction that allows for use of prior knowledge via penalty functions with data from multiple coils.
-
Combined parallel imaging and compressed sensing reconstruction for accelerated perfusion MRI.
-
Low-rank plus sparse reconstruction for highly accelerated dynamic MRI.
An accurate, rapidly converging, low-compute algorithm for approximating solutions to the least absolute shrinkage and selection operator (LASSO) problem.
A method for parameter mapping in accelerated MRI.
Scripts for two-dimensional modeling of diffusion.
A MATLAB and Julia toolbox for reconstruction of quantitative maps from MR fingerprinting data.
A magnetic resonance fingerprinting method for tractography.
-
A convolutional neural network for anatomically-guided PET reconstruction.
-
Respiratory motion compensation, automatic bolus timing, and coil-weighted unstreaking for GRASP.
A highly accurate radial reconstruction for diffusion spectrum imaging.
Scripts for modeling the time-varying diffusion coefficient in biological tissue.
Total generalized variation for denoising and reconstruction of accelerated MR data.
A non-uniform fast Fourier transform with Kaiser-Bessel gridding for machine learning applications in PyTorch.
-
A reconstruction method that sorts dynamic data into motion states in accelerated MRI.
Radiofrequency and Dielectric Tools
A radio-frequency coil inductance and capacitor calculator for Android.
Software for calculating specific absorption rate.
-
Scripts for temperature simulations.
An open-source, web-based, cross-platform inductance and capacitor calculator for building birdcage RF coils.
-
An online calculator for creating dielectric MRI phantoms.
-
Software for predicting the effect of electromagnetic fields on MR images.
-
MATLAB software for calculating conductivity and permittivity of biological tissues.
-
MATLAB code for calculating local average specific absorption rate for a given mass of tissue.
MATLAB code for determining the UISNR at the center of a dielectric sphere.
Image Analysis Software
Medical image analysis software.
-
Code for low-rank plus sparse decomposition of orientation distribution functions.
Software for robust standard-model parameter estimation from diffusion MRI data.
Other Tools
Scripts for calculating diffusion kurtosis imaging metrics.
-
Fast and robust temperature maps from raw Siemens MRI data on OSX.
-
Fast and robust temperature maps from raw Siemens MRI data on PC.
-
Data and code for precise, sample-based modeling of axon microgeometry.
MATLAB scripts for learned pulse sequence parameters of magnetization-prepared gradient echo sequences used in multi-component T2 and T1ρ mapping.
-
Compare machine learning performance across classifiers and datasets.
MATLAB scripts for data-driven optimization of magnetization-prepared gradient echo sequences used in T1ρ mapping.
-
Graphical user interface for designing two-dimensional RF pulse sequences for parallel transmit systems in MATLAB.
-
Graphical user interface for parallel transmission radio-frequency shimming in MATLAB.
Note
Software downloads and other resources made available by CAI2R are provided free of charge and come without any warranty. Neither CAI²R nor NYU Grossman School of Medicine take any liability for potential problems or damage of any kind resulting from the use of the files provided. Operation of the software is at the user’s own risk. The resources provided are not medical products and must not be used for making diagnostic decisions.
Some of the resources provided by CAI2R are works in progress. Therefore, some functionality may be missing and we cannot guarantee that the software is completely free of errors. Please check regularly for updated versions of the available download items (many resource pages contain version-history information).