Developer position: Machine learning for Brain Neuroimaging

KEY WORDS

CITY

COUNTRY

France

DETAILS OF THE OFFER

Working place:  "Mind" research team (Inria Saclay and CEA)

Missions:

The job proposal includes two main, equally important, missions. The first mission is to set up robust pre-processing pipelines for data coming from our collaborating institutions such that we can transfer our scientific advancements to clinical and pre-clinical settings. The users of these pipelines will be the members of the MIND team as well as close collaborators. These pipelines will be concerned mainly with two different image modalities, functional and diffusion MRI. Data from all these modalities has to first undergo artifact correction, denoising, and standardization. The pipelines addressing previously described use cases will build on tools such as Fmriprep, Nilearn, Dipy, ANTs and FreeSurfer. Beyond basic use of these powerful tools, we need to add steps for quality control of the output or special adaptations to the pathological cases in our collaboration, such as brain tumors. A second, equally important objective is to ensure the development of Nilearn. In relation to the other goal, a planed development is an improvement of the connection with other tools from the ecosystem, in particular BIDS standards: leveraging these standards will facilitate the analysis of BIDS-organized datasets. In general, we would like to automate more the main analytical steps in Nilearn to improve user’s experience. We also want to upgrade interactive visualization tools with the integration of NiiVue. Finally, depending on the developer’s skills on optimization, we aim at improving performance of some Nilearn functionality based on profiling/memory management analysis, use of GPU computing etc.

More in detail, the following actions will be undertaken:

• Organization of several datasets available to the MIND team using BIDS conventions
• Further automate Nilearn pipelining in BIDS contexts to improve user experience
• Setting up initial Functional and Diffusion MRI pipelines based on existing processing tools
• Extraction of High Level Features from the preprocessed dataset for use in neuroscientifi studies
• Integration of Niivue for interactive visualization in Nilearn
• Improve performance of some Nilearn functionality (profiling/memory management/GPU...).
• Animation of the Nilearn community

Qualifications:

• Love high-quality code and open source
• Worry about users and like to communicate
• Be curious about data (ie like looking at data and understanding it)
• Have an affinity for problem-solving tradeoffs
• Good scientific Python coders
• Enjoy interacting with a community of developers
• Interest in brain imaging and its applications.
• Experience in optimization is a plus

+info

TYPE OF JOB

TYPE OF CONTRACT

CDD (Temporary / 2 years )

REMUNERATION

EMPLOYMENT START DATE