Hi all, this is my first post since I got accepted to GSoC 2015. I am really excited about the start of the coding period and about being part of the greater community of the Python Software Foundation. Honestly, I am a bit scared, but I like the challenge and I am working with the best people. The Dipy team is really great! During the community bonding period I was able to interact with some of my mentors and draw a general plan of the coding phase.
First, a short intro to my project which is called: "Tissue classification to improve tractography."
This is the abstract of my project:
As stated in my initial proposal, the first task for the the community bonding period was to read and discuss the paper by Zhang et al, 2001 (Yongyue Zhang; Brady, M.; Smith, S., "Segmentation of brain MR images through a hidden Markov random field model and the expectation-maximization algorithm," Medical Imaging, IEEE Transactions on , vol.20, no.1, pp.45,57, Jan 2001). This paper gives us a closer idea of how to approach the segmentation algorithm for MRI T1-weighted images of the brain. The goal is to derive partial volume estimates (PVEs) for each of the tissues and compartments of the brain, i.e. grey matter, white matter, and cerebrospinal fluid. With the mentors we defined the main strategy to code the segmentation algorithm proposed in the paper, which parts of the theory we would like to implement and which ones not as well as the general assumptions about the inputs to the program.
First, a short intro to my project which is called: "Tissue classification to improve tractography."
This is the abstract of my project:
- Diffusion Magnetic Resonance Imaging (dMRI) is used primarily for creating visual representations of the structural connectivity of the brain also known as tractography. Research has shown that using a tissue classifier can be of great benefit to create more accurate representations of the underlying connections. The goal of this project is to generate tissue classifiers using dMRI or a different MRI modality e.g. T1-weighted MRI (T1). This reduces to an image segmentation task. I will have to implement popular segmentation algorithms using T1 and invent a new one using dMRI data.
As stated in my initial proposal, the first task for the the community bonding period was to read and discuss the paper by Zhang et al, 2001 (Yongyue Zhang; Brady, M.; Smith, S., "Segmentation of brain MR images through a hidden Markov random field model and the expectation-maximization algorithm," Medical Imaging, IEEE Transactions on , vol.20, no.1, pp.45,57, Jan 2001). This paper gives us a closer idea of how to approach the segmentation algorithm for MRI T1-weighted images of the brain. The goal is to derive partial volume estimates (PVEs) for each of the tissues and compartments of the brain, i.e. grey matter, white matter, and cerebrospinal fluid. With the mentors we defined the main strategy to code the segmentation algorithm proposed in the paper, which parts of the theory we would like to implement and which ones not as well as the general assumptions about the inputs to the program.
No comments:
Post a Comment