CT scanning of dinosaur fossils

CT scanning is the "X-ray vision" of paleontology. It captures complete 3D internal data without damaging fossils, revealing scientific secrets inaccessible to traditional methods. From brain cavity morphology to bone microstructure, CT scanning is reshaping our understanding of dinosaur biology.

Device selection depends on specimen size and research goals. Medical CT is suitable for large specimens like skulls, offering a resolution of 0.3-0.5 mm with short scan times but limited detail. Industrial micro-CT achieves resolutions down to 5-50 µm, ideal for fine specimens such as teeth, small bones, and eggshells, though it requires longer scan times and has sample size limitations. Synchrotron radiation CT is the ultimate tool, providing sub-micron resolution and high penetration power, but access requires scheduling time at a major scientific facility.

The data processing workflow includes reconstructing raw projection data into a sequence of tomographic slices, segmenting distinct tissues and matrix, and performing 3D visualization and quantitative analysis. Segmentation is the most time-consuming step, requiring manual or semi-automatic differentiation of bone, teeth, voids, and host rock. While deep learning segmentation algorithms have significantly improved efficiency in recent years, expert validation remains necessary. The resulting 3D models enable downstream analyses such as geometric morphometrics, finite element analysis, and fluid dynamics simulations.

Key breakthroughs include: reconstructing dinosaurs' sensory capabilities and intelligence via brain cavity endocasts; estimating individual age and growth rates from tooth growth lines; inferring hatching environments through eggshell pore structures; and analyzing vocalization mechanisms via sinus systems. Open sharing of CT data is accelerating these discoveries. We recommend that research institutions establish standardized CT data acquisition and archiving protocols to ensure long-term data availability and comparability.