Computational Core¶
computational-core describes software projects authored by John J. Lee and collaborators for the Computational Core of the Neuroimaging Laboratories at Washington University in St. Louis. These projects support research programs of the organizational units aforementioned.
While the scope of projects is diverse, research themes most commonly involve:
biophysical models of brain metabolism and function
instrumentation and data specific for positron emission tomography
instrumentation and data specific for resting-state functional magnetic resonance imaging
instrumentation and data specific for intracranial electroencephalography
data archives based on XNAT
standardized data formats such as 4dfp, NIfTI, CIFTI, and BIDS
inferential methodologies drawn from Bayesian statistics, expectation maximization, Markov chain Monte Carlo, graphical models, and deep learning
reuse of mature pre-existing projects such as FSL, FreeSurfer, Tensorflow, Pytorch, and MONAI.
This project is under active development.
Contents¶
package mlfourd
transparently supports 4dfp, NIfTI, CIFTI, FreeSurfer, iEEG and BIDS data formats
provides adapter patterns implementing client interfaces familiar for neuroscience tasks
uses state patterns instantiating lightweight objects optimized for categories of data and behavior
package mlraichle
provides biophysical models for oxygen and glucose metabolism
provides image-derived input functions
supports instrumentation related to the Siemens Biograph mMR
package mlvg
supports instrumentation related to the Siemens Biograph Vision
package mlarbelaez
supports instrumentation related to the Siemens ECAT EXACT HR+
project cc-graph-nets
implements Deep Minds’ Graph Nets library for neuroimaging
project cc-trax
implements Google Brain’s Trax and transformers for neuroimaging
project cc-vision-transformer
implements vision transformers and MLP-mixer architectures for neuroimaging
Feature highlights
API Reference