New! Engineer Position-MeCA research group

Extraction of sulci from fetal brain MRI data

Duration : 12 months, renewable

Level : Master 2, Engineering school, PhD 

Location : Institut de Neurosciences de la Timone, 13005, Marseille

Team: Methods and Computational Anatomy (MeCA research group)

Supervision : Guillaume Auzias, guillaume.auzias@univ-amu.fr

Context: The human cerebral cortex undergoes dynamic and regionally heterogeneous development during gestation[1, 2]. For more than a decade, the MeCA team has collaborated with the hospital of La Timone in Marseille to build and process a unique large database of cerebral MRI acquired in utero in clinical routine. We also developed an image processing pipeline dedicated to segment the different brain tissues from the initial MRI acquired in clinical routine at the Hospital of La Timone (https://fetpype.github.io/fetpype/). This pipeline has been used to analyse the evolution of the volume of brain structures with age in [1]. We now aim at extending this type of analysis to specific measures extracted from the cortical sulci as illustrated in Fig1. To do so, we need to improve our image processing pipeline by adapting the approach introduced in  [3] for extracting the sulci. 

Figure 1: Left: MRI scan of a fetal brain. Center: tissues (white matter, and cortex) segmentation. Right: example of the result of the sulci extraction we would like to implement for fetal data. This image from [2] has been obtained from a post-natal MRI acquired on a preterm baby.

The recruited Engineer will contribute to this line of research by designing and implementing various image processing tools, as well as applying the pipeline to large samples of data.

[1] A. Mihailov, et al, Communications Biology, (2025); https://doi.org/10.1038/s42003-025-08155-z   

[2] H. de Vareilles et al., Developmental Cognitive Neuroscience (2023); https://doi.org/10.1016/j.dcn.2023.101249 [3] J.-F. Mangin et al., NeuroImage (2004);  https://doi.org/10.1016/j.neuroimage.2004.07.019.

Main Tasks: The Engineer will be in charge of: 

• Adaptation of the process for extracting the sulcal graphs to fetal MRI data.
• Develop and test various improvements to our fetal MRI processing pipeline, which consists of several steps such as super-resolution of a 3D high resolution volume from the acquired stacks of 2D MRI, fine-tuning brain extraction deep learning models, improving the tissue segmentation algorithm, and testing new techniques from the literature.
• Adaptation of the segmentation techniques to pathological cases such as corpus callosum agenesis
• Process large samples of data (extraction of shape descriptors, quality control, descriptive statistics)
• Software development (https://brain-slam.github.io/slam/  and https://fetpype.github.io/fetpype/)

Expected Skills:

• Knowledge:
  • Proficient  in Python for image processing
  • Proficient in Git, GitHub and Linux based environment
• Skills:
  • Organizational skills (planning and project management, rigor in work organization, manage deadlines).
  •  Capacity and desire to acquire new skills.
• Soft Skills:
  •  Good interpersonal and communication skills will be appreciated (ability to work alone and in a team).

Work Environment: The Engineer will join the MeCA research group at the Institut des Neurosciences de la Timone, in Marseille, France. The Meca team combines expertise in processing of large fetal MRI databases and surface based morphometry methods. 

How to apply?

Applications from individuals of all genders are encouraged. 

Please send your CV + motivation letter to guillaume.auzias@univ-amu.fr