Modeling textures in mammograms by anisotropic fractal random fields
A Master internship position in image processing will open in 2027 in the Signal, Image and Sounds (SIMS) team in LS2N in the framework of the UNGRAM project led by Barbara PASCAL (CNRS researcher, bpascal-fr.github.io).
Objectives
The overarching goal of this internship is to model the textures seen in mammograms by anisotropic fractal random fields in order to precisely characterize the different tissues depending on their fractal attributes. In the end, these interpretable texture features are to be used to evaluate the breast cancer risk of a women based on the subtle patterns observed in a single mammogram.
Methodology
This subject will combine mathematical tools enabling the modeling of scale-invariant textures by Gaussian random fields, synthesis algorithms to generate such fields with various characteristic features, and statistical procedures to fine-tune the model parameters so that synthetic textures matches as closely as possible real mammographic textures.
A detailed description can be found here: https://uncloud.univ-nantes.fr/index.php/s/QPFDes4HAygGeoi.
Application
To apply, please send a CV, transcripts of License and Master grades, support letter(s) with references and a paragraph of motivations to barbara.pascal@cnrs.fr.
