Implementation of soma and neurite density imaging (SANDI) in the accelerated microstructure imaging via convex optimization (AMICO) framework
Title: Implementation of soma and neurite density imaging (SANDI) in the accelerated microstructure imaging via convex optimization (AMICO) framework
Project lead: Marco Palombo (email: marco.palombo@ucl.ac.uk; Twitter: @MarcoPalombo3)
Project collaborators: Simona Schiavi (email: simona.schiavi@univr.it; Twitter: @simonaschiavi24) Alessandro Daducci (email: alessandro.daducci@univr.it; Twitter: @ADaducci)
Registered Brainhack Global 2020 Event: Brainhack Atlantis The Atlantic Ocean - Micro2Macro
Project Description:
AIM
The aim of the project is to implement an algorithm for fast and robust fitting of complex microstructure imaging methods, such as the soma and neurite density imaging (SANDI), using the accelerated microstructure imaging via convex optimization (AMICO) framework.
What is SANDI?
SANDI (https://doi.org/10.1016/j.neuroimage.2020.116835) is a novel imaging technique based on diffusion MRI and biophysical modelling, designed to provide maps of MR indices of apparent soma density and size, as well as apparent neurite density. As such, SANDI is of interest for studies involving the characterization of brain cytoarchitectonics in a wide range of conditions: e.g., brain development, aging, plasticity, neuroinflammation, neurodegeneration, and more.
What is AMICO?
The AMICO framework (https://doi.org/10.1016/j.neuroimage.2014.10.026) offers the possibility to re-formulates virtually any microstructure imaging model as convenient linear systems which, then, can be efficiently solved using very fast algorithms. To date, it has been successfully employed with other models, such as NODDI and ActiveAx. We will implement an AMICO version of SANDI building upon the code publicly available at https://github.com/daducci/AMICO
Why SANDI-AMICO
The fitting of the SANDI model originally proposed is based on machine learning regression, which provides ultra-fast model parameters estimation but requires the careful design of both the training strategy and the machine learning model.
The AMICO framework can overcome these issues and provide the scientific community with a more agnostic algorithm for the fast and robust estimation of SANDI indices.
Data to use:
We will use unpublished data collected at high field and high diffusion gradient strength on the mouse brain. Since it is unpublished, the data will be made available privately by the project leader.
Link to project repository/sources: https://github.com/daducci/AMICO Palombo, Marco, et al. “SANDI: a compartment-based model for non-invasive apparent soma and neurite imaging by diffusion MRI.” NeuroImage (2020): 116835. (https://doi.org/10.1016/j.neuroimage.2020.116835) Daducci, Alessandro, et al. “Accelerated microstructure imaging via convex optimization (AMICO) from diffusion MRI data.” NeuroImage 105 (2015): 32-44. (https://doi.org/10.1016/j.neuroimage.2014.10.026)
Goals for Brainhack Global 2020: Deliverable 1: stable version of SANDI-AMICO; Deliverable 2: tested version of SANDI-AMICO on one suitable dataset;
Good first issues:
- Implementing data and protocol loader: read NIFTI file; read b values and directions; check data consistency; separate data in shells and average across directions;
- Implementing database generator: define functions to generate response functions to be used in the AMICO linear system according to the SANDI model and the Gaussian Phase Distribution approximation;
- Implementing the optimizer: define the objective function, setup the optimizer, perform the voxelwise fitting;
- Implementing the result writer: save the SANDI maps in NIFTI format, keeping the correct header file;
- Test performances: generate known ground-truth conditions using analytical simulations and different noise characteristics (e.g. Gaussian, Gaussian + floor; Rician);
- Write up an initial tutorial on the wiki page of the AMICO repository.
Skills: Python programming; Familiar with DWI data processing (e.g. NIFTI/DICOM format, diffusion-weighted MRI protocols etc.); Basic understanding on inverse linear problems (least square method, common regularization techniques, etc.);
Tools/Software/Methods to Use: Python 3, AMICO (pip install dmri-amico), Camino, visual studio code or spider or jupyter.
Communication channels: https://mattermost.brainhack.org/brainhack/channels/micro2macro-sandi_in_amico
https://zoom.us/ (link will be posted in mattermost channel)
Project labels #dMRI, #modelling, #AMICO, #SANDI, #Microstructure, #InVivoMicroscopy, #Neuro
Type of project: #coding_methods #pipeline_development
Project development status: #1_basic structure
Topic of the projet: #diffusion #MR_methodologies
Tools used in the project: #DIPY, #FSL, #Jupyter, #AMICO #Camino
Tools skill level required to enter the project (more than one possible): #familiar
Programming language used in the project: #Python
Modalities involved in the project (if any): #DWI, #MRI
Git skills reuired to enter the project (more than one possible): #0_no_git_skills
Project Submission
Submission checklist
Once the issue is submitted, please check items in this list as you add under ‘Additional project info’
- I added all of the labels I want an associate to my project
Project Submission
Submission checklist
Once the issue is submitted, please check items in this list as you add under ‘Additional project info’
- Link to your project: could be a code repository, a shared document, etc.
- Goals for Brainhack Global 2020: describe what you want to achieve during this brainhack.
- Flesh out at least 2 “good first issues”: those are tasks that do not require any prior knowledge about your project, could be defined as issues in a GitHub repository, or in a shared document.
- Skills: list skills that would be particularly suitable for your project. We ask you to include at least one non-coding skill. Use the issue labels for this purpose.
- Chat channel: A link to a chat channel that will be used during the Brainhack Global 2020 event. This can be an existing channel or a new one. We recommend using the Brainhack space on Mattermost.
Optionally, you can also include information about:
- Number of participants required.
- Twitter-sized summary of your project pitch.
- Provide an image of your project for the Brainhack Global 2020 website.
We would like to think about how you will credit and onboard new members to your project. If you’d like to share your thoughts with future project participants, you can include information about:
- Specify how you will acknowledge contributions (e.g. listing members on a contributing page).
- Provide links to onboarding documents if you have some: