Rachelle Saint-Fort, MS

3D Computational Gut Cell Characterization Across Animal Models

Rachelle Saint-Fort, Khai C. Ang, Mee S. Ngu, Daniel J. Vanselow & Keith C. Cheng Department of Pathology, The Jake Gittlen Laboratories for Cancer Research, Penn State College of Medicine, Hershey PA

Emerging computational tools are revolutionizing 3D anatomical image analysis, enabling high-throughput pipelines for faster, more accurate diagnoses and reducing clinical biases. However, identifying cellular morphological changes linked to mutations or diseases remains challenging, particularly given the need to characterize diverse cell types across animal models. Our lab has optimized a microCT-based histological technique, termed histotomography, which offers histology-level resolution and 3D microscopy to discern subtle phenotypes essential for diagnosis.   We are developing a quantitative approach to characterize morphological variations in normal cells, leveraging mathematical parameters to capture volumetric and morphometric features, along with intercellular relationships. As proof-of-concept, we are prioritizing the study of gut epithelial cells, chosen for their well-documented structural alterations in gastrointestinal disorders. Using our reference atlases, we are defining conserved phenotypic features in Daphnia magna to recognize analogous cell types in wild-type zebrafish and Mexican axolotl larvae scans. Using a synchrotron-based system with a 5mm field-of-view / 0.5-micron resolution, shape statistics (including volume, shape, and spatial relationships) derived from segmented cells are used to define distributions of normal phenotypic variations. This phenotypic data will guide the development of a machine learning algorithm for automated cell type recognition, accounting for biological variability across datasets. Ultimately, this approach supports unbiased, whole-organism computational phenotyping, offering a robust framework for disease modeling and diagnosis.