Statusvortrag im Promotionsver­fahren von Frau Roua Rouatbi

Titel: "Quantifying and comparing dynamic biological shapes“

Abstract:
The human visual system excels at recognizing and grouping objects by shape, forming equivalence classes that are invariant under shape-preserving transformations such as rotation, translation, reflection, and scaling. Developing quantitative metrics that capture this notion of shape similarity, however, remains a central challenge in biomedical imaging.
In my status talk, I will present the Push-Forward Signed Distance Morphometric (PF-SDM), a framework for shape quantification in biomedical imaging. The PF-SDM provides a compact and mathematically smooth representation of closed shapes, encoding both geometric and topological properties, including their topological skeletal structure and rotational symmetries.
The framework naturally extends to time-dependent shapes and supports the fusion of spatial intensity information such as genetic or molecular markers with shape dynamics.
I will introduce the theoretical foundations of PF-SDM, benchmark its performance on synthetic two- and three-dimensional datasets, and demonstrate its utility in addressing biological questions, including the quantification and comparison of developmental trajectories in gastruloids.