Events and News

Methods for a complex MRI-based model of the brain

News
Motivation.
At present, the construction of brain models, including the detection of brain pathologies, is based on several types of examinations:
  • Magnetic resonance (MR) morphometry -- allows us to identify abnormalities in the structure of the brain, based on an assessment of the shape, volume, and structure of its individual regions.
  • Tractography allows us to detect abnormalities in the brain's conductive pathways (white matter tracts), examine their directions, displacements or deformations, and evaluate the connectivity of brain regions.
  • Functional magnetic resonance imaging (fMRI) -- aimed at analyzing the neural activity of the brain or spinal cord, based on changes in the relative concentrations of oxygen-rich and oxygen-poor blood in areas of activation; allows in some cases to link processes occurring in the body (or cognitive tasks performed by the patient) to the activity of groups (networks) of neurons

There are algorithms and software tools for processing the data (obtained from tomographs) for each of the above examinations. These examinations are performed independently, so it is impossible to build a comprehensive model of the brain.

Project Objectives:
  • Exploring the possibilities of MR-morphometry, tractography, fMRI to build models of the brain and detect pathologies
  • Application of artificial intelligence methods to automate MRI data analysis tasks
  • Development of approaches and methods for building a comprehensive brain model based on MR-morphometry, tractography, fMRI

Existing software systems:
  • https://web.conn-toolbox.org - resting fMRI processing and analysis (CONN)
  • https://surfer.nmr.mgh.harvard.edu - morphometry
  • http://dsi-studio.labsolver.org - software for tractography (+stat analysis)
  • http://www.neuro.uni-jena.de/cat/ - CAT12 toolbox for morphometry
  • https://www.fil.ion.ucl.ac.uk/spm/software/spm12/ - matlab plugin (SPM) under which CONN works (functional connectivity toolbox)
  • https://lcni.uoregon.edu/downloads/mriconvert/mriconvert-and-mcverter - dicom converter into required format

Jobs:
  • Research Engineer
  • Data Analysis Engineer
  • Machine Learning Engineer
  • Programmer

Additional Information:
  • According to the results of the project it is possible to write and defend Bachelor's, Master's and PhD works
  • Payment is based on interview results
  • Perhaps training the necessary skills in the course of the project

Contacts:
To participate in the project, send an application to job@etu.ai