Axon-based Parcellation of the Corpus Callosum in Subjects with Multiple Sclerosis


Multiple sclerosis (MS) is a disease affecting approximately 1 out of 1000 people in the U.S. Patients suffer from a variety of neurologic symptoms. The disease typically manifests itself in people in their 20s and 30s, and progresses over years and decades, often severely limiting the independence and productivity of those affected. The corpus callosum (CC) is part of the brain, it connects left and right hemispheres. This area shrinks in patients with MS. 


In order to analyze the CC in Magnetic Resonance Images (MRI) (Aribandi, M., 2015), it is necessary to take into account the distribution and shape of axons (connections) traversing it (Aboitiz et al., 1992). Although basic parcellations of the CC have been defined in previous works (Hofer and Frahm, 2006; Hofer et al., 2015; Witelson et al. 1989), a more flexible distribution, adaptable to different shapes of CC, could enhance the analysis of the CC. The objective of this project is to propose an automatic parcellation of the CC where each region is adapted to the distribution of axons inside the CC. 


  1. Implement a robust extraction of the middle-line of the CC
  2. Analyse the Intensity distribution of the CC along the middle-line
  3. Define a method for region selection (parcellation) based on the intensity distribution in the CC
  4. Correlate region-intensities and clinical data in a group of MS subjects
  5. Improve and refine the method


  What skills do you need?

  • Programming skills in C++, Matlab, and python (javascript is a plus)
  • Image processing background. Knowledge of ITK, VTK is a plus.

  What skills you will acquire?

  • Improve presentation skills.
    • You will attend monthly, sometimes weekly, presentations from researchers of groups collaborating with us. 
    • You will have to present to the group at least three times (first: project and objectives, fourth: advances, and last month:preparation for defense)
  • Team working
    • Expose your ideas within a multidisciplinary team
  • Data management
    • You will be exposed to different types of data, typically images (MRI mostly) and clinical data. You will learn the best practices to handle this information.
  • Networking
    • Being in a multidisciplinary team and linked to research groups all around the world you will enhance your networking skills. ….

  What you will learn?

  • Brain
    • You will learn about brain morphology and physiology
  • Imaging
    • You will understand the basics of MRI and how to interpret them, with emphasis on the corpus callosum and white matter.
    • You will handle Diffusion Tensor Images (DTI) and will understand the basic physics underneath this type of acquisition.


Team / environment

  • The Center for Neurological Imaging (CNI) team involves physicians, computer scientists, and virtual reality specialists. As we collaborate with many groups you will also communicate with statisticians, senior software developers, and researcher from multiple diverse fields.


  1. Aboitiz, F., Scheibel, A. B., Fisher, R. S. & Zaidel, E., 1992. Fiber composition of the human corpus callosum. Brain Research, pp. 143-153.
  2. Aribandi, M., 2015. Imaging in Agenesis of the Corpus Callosum. MedScape. 
  3. Granberg, T. et al., 2015. Corpus callosum atrophy is strongly associated with cognitive impairment in multiple sclerosis: Results of a 17-year longitudinal study. Multiple Sclerosis, pp. 1151-1158. 
  4. Hofer, S. & Frahm, J., 2006. Topography of the human corpus callosum revisited–comprehensive fiber tractography using diffusion tensor magnetic resonance imaging.. NeuroImage, pp. 989-994. 
  5. Hofer, S., Wang, X., Roeloffs, V. & Frahm, J., 2015. Single-shot T1 mapping of the corpus callosum: a rapid characterization of the fiber bundle anatomy. Frontiers in Neuroanatomy, Volume 9, p. 57. 
  6. Witelson, S. F., 1989. Hand and sex differences in the isthmus and geni of the human corpus callosum. A postmortem morphological study. Brain, 112 ( Pt 3):799-835.

See also: Project