Register now After registration you will be able to apply for this opportunity online.
This opportunity is not published. No applications will be accepted.
Deep learning techniques for quantitative characterization of microscopy multidimensional images
Segmentation of cells/structures in 3D microscopy images
Keywords: image segmentation; deep learning; 3D microscopy
Blood cell production mainly occurs in the cavities of long bones (bone marrow) in a complex and spatially heterogenous process. With new imaging methodologies, it has been possible to study how the function of some cells is regulated by the spatial context. We have developed a strategy to create multiscale 3D images of different bone marrow cellular components with unprecedented resolution. With vast amounts of data already available, the challenge resides in the generation of tools for automatic image segmentation and analysis.
Blood cell production mainly occurs in the cavities of long bones (bone marrow) in a complex and spatially heterogenous process. With new imaging methodologies, it has been possible to study how the function of some cells is regulated by the spatial context. We have developed a strategy to create multiscale 3D images of different bone marrow cellular components with unprecedented resolution. With vast amounts of data already available, the challenge resides in the generation of tools for automatic image segmentation and analysis.
In this project, different state-of-the-art deep learning methodologies will be explored to overcome the image segmentation challenges posed by very high-resolution microcopy datasets, which do not yield themselves easily to off-the-shelf image processing algorithms. The resultant segmentation of the different bone marrow components is important to better understand how these structures affect blood production.
In this project, different state-of-the-art deep learning methodologies will be explored to overcome the image segmentation challenges posed by very high-resolution microcopy datasets, which do not yield themselves easily to off-the-shelf image processing algorithms. The resultant segmentation of the different bone marrow components is important to better understand how these structures affect blood production.