Our results demonstrate that high‑resolution μCT coupled with deep‑learning segmentation can recover previously inaccessible soft‑tissue anatomy across a wide taxonomic breadth. The pipeline reduces processing time per specimen from , thereby scaling paleontological investigations to dataset sizes previously unmanageable.
Bayesian phylogenies recovered (Fig. 3). Lineage‑through‑time (LTT) analysis identified three diversification pulses: ginasavagex cambro
From each mesh, 45 geometric and topological traits were quantified (e.g., body volume, appendage curvature, canal diameter). Trait values were standardized (z‑score) and subjected to to construct a morphospace. ginasavagex cambro