A precise atlas of the brain is an essential step to uncover brain functional mechanisms in the healthy and diseased brain. Non-invasive neuroimaging have provided reproducible parcellations of the human brain. In animal models of neuroscience, parcellations have usually relied on cytoarchitecture (changes in cellular density over space) to infer on parcellation. We propose to implement homologous parcellation methods derived from human fMRI data to develop a mouse brain atlas using a collection of >1000 high-resolution mouse fMRI data previously acquired using dedicated MRI systems.
Specific goal
we propose to implement the 'Gordon' method (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4677978/) to brain parcellation to mouse fMRI data preprocessed and normalized into a common space (AIBS ccfv3). The mouse fMRI data is a collection from different dataset published since 2014 on the central.xnat.org and openneuro.org repository. The data is available in readily preprocess format
The outcome is a 3D volumetric mouse brain atlas, and a comparison to its cytoarchitecture counterpart, and to human atlases.
Skills required to participate
An interest in neuroscience.
Coding skills
Python or matlab. The original method was carried out using the watershed function in matlab. Such a function should exist in python too.
language
level of expertise required
python
beginner
matlab
beginner
software specific sills
no software specific skills required
GIT / GITHUB skills required
no software specific sills required |
Integration
The project is a basic science project that can interest a range of people interested in neuroscience, including comparative neuroanatomy, physiology, imaging, and computational neuroscience.
Functional parcellation of the mouse brain
Project Description
A precise atlas of the brain is an essential step to uncover brain functional mechanisms in the healthy and diseased brain. Non-invasive neuroimaging have provided reproducible parcellations of the human brain. In animal models of neuroscience, parcellations have usually relied on cytoarchitecture (changes in cellular density over space) to infer on parcellation. We propose to implement homologous parcellation methods derived from human fMRI data to develop a mouse brain atlas using a collection of >1000 high-resolution mouse fMRI data previously acquired using dedicated MRI systems.
Specific goal
we propose to implement the 'Gordon' method (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4677978/) to brain parcellation to mouse fMRI data preprocessed and normalized into a common space (AIBS ccfv3). The mouse fMRI data is a collection from different dataset published since 2014 on the central.xnat.org and openneuro.org repository. The data is available in readily preprocess format
The outcome is a 3D volumetric mouse brain atlas, and a comparison to its cytoarchitecture counterpart, and to human atlases.
Skills required to participate
An interest in neuroscience.
Coding skills
Python or matlab. The original method was carried out using the watershed function in matlab. Such a function should exist in python too.
software specific sills
GIT / GITHUB skills required
Integration
The project is a basic science project that can interest a range of people interested in neuroscience, including comparative neuroanatomy, physiology, imaging, and computational neuroscience.
Preparation material
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4677978/ : the original method described. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3174820/ : an alternative method. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6596781/ : an exemple of mouse fMRI data being implemented and compared to neuroanatomy
Link to your GitHub repo
https://grandjeanlab.github.io/
Communication
To be determined. I intended to attend the lectures rather than hackathon. Reach me at: https://grandjeanlab.github.io/pages/contact.html