Tutorial: Medical image analysis for structural and functional MRI

Brandon Whitcher, GlaxoSmithKline, Clinical Imaging Centre, Hammersmith Hospital, United Kingdom
Pierre Lafaye de Micheaux, Department of Mathematics and Statistics, Universite de Montreal, Canada
Bradley Buchsbaum, Rotman Research Institute, Canada
Jörg Polzehl, Stochastic Algorithms and Nonparametric Statistics, Weierstrauss Institute for Applied Analysis and Stochastics, Germany


The field of medical imaging covers a vast range of disciplines and applications. There is a growing collection of open-source software (OSS) solutions for all aspects of data management, processing, analysis and visualization. This tutorial will provide four specific R packages from the Medical Imaging Task View and apply them to structural and functional MRI data.


By the end of the tutorial attendees will be able to:


  1. Data Import/Export using AnalyzeFMRI by Pierre Lafaye de Micheaux
    1. Data import/export using the Analyze format
    2. Data import/export using the NIFTI format
    3. Conversion from Analyze to NIFTI
    4. From voxel indices to spatial coordinates
    5. Quaternions, rotations and the like
    6. Visualization of images using the GUI
    7. Spatial/temporal ICA using the GUI
  2. Functional MRI using Neuroimage by Bradley Buchsbaum
    1. Data Structures for 3D and 4D images
    2. Statistical modelling of fMRI data
  3. Diffusion tensor imaging using dti by Jörg Polzehl
    1. Diffusion tensor model and derived characteristics
    2. Q-ball imaging
    3. Tensor mixture models
  4. Dynamic Contrast-enhanced MRI using dcemri by Brandon Whitcher
    1. Introduction to DCE-MRI and data visualization
    2. T1 estimation from multiple flip angles
    3. Estimating gadolinium concentration
    4. AIF extraction and modelling
    5. Kinetic parameter estimation
  5. Overview of the Medical Imaging Task View and general discussion

Potential attendees

Statisticians, medical physicists and researchers with an interest in neuroscience and/or oncology are encouraged to attend.


Attendees should have a basic understanding of an interpreted programming language, such as R (preferred) or Matlab. Attendees should also have a basic understanding of statistical methodology, such as summary statistics, hypothesis tests, linear regression, non-linear regression, etc. Basic knowledge of medical imaging (specifically MRI) is an advantage but not necessary.

Tutorial Materials

Materials are here, here and here.