Reconstructing a Fossil Fish Braincase with Dragonfly 3D World

November 12, 2025

Understanding the internal skeletal anatomy of fossilized organisms has long been a challenge in paleontology. The braincase, often concealed beneath layers of dermal bone, is rarely visible without damaging the specimen. Traditional mechanical preparation cannot expose delicate internal structures without loss of information.

Researchers have now overcome this challenge by using high-resolution micro-computed tomography (µ-CT) and advanced visualization tools in Dragonfly 3D World to study the braincase of Lepidotes, a bony fish that lived approximately 180 million years ago in ancient European seas.

Fossil bony fish Lepidotes from the Lower Jurassic of Germany, tail fin missing. Specimen acid prepared and dusted with ammonium chloride for photography.

Why It Matters

The study of the Lepidotes braincase offers new insights into the evolution and morphology of fossil fishes. The skull of Lepidotes is characterized by robust dermal bones covering most of the head, leaving only the eye socket and mouth open. Beneath these protective plates lies an ossified braincase - a structure that, until now, was inaccessible for comprehensive detailed analysis.

Using µ-CT imaging, researchers were able to visualize and segment these internal ossifications without destructive mechanical preparation. The project highlights how digital imaging and 3D World’s segmentation capabilities are improving paleontological studies by revealing structures that were once hidden from view.

Skull of Lepidotes from the Lower Jurassic of Germany. Photo of specimen dusted with ammonium chloride and CT 3D view. Some disassembled bony elements of the braincase are partly visible in the eye socket.

Testing and Validation

A three-dimensionally preserved Lepidotes specimen was scanned using a Nikon XT H 320 system at 215 kV X-ray voltage, 240 µA current, with a 1.5 mm Cu filter and 77.834 µm voxel size.

The µ-CT data was visualized and analyzed using Dragonfly 3D World (version 2024.1). The software’s intuitive tools, detailed documentation, and video tutorials simplified complex segmentation tasks—critical for distinguishing delicate endocranial structures embedded beneath denser dermal bone layers.

Combination of 3D and 2D views. Upper left is a right lateral 3D view of the Lepidotes skull. Grayscaled bone material of higher density partly conceals our ROIs below the skull roof. Complete visibility and segmentation of the ROIs were achieved with additional three 2D sectional views from different interdependent angles.

Observations and Results

Endocranial bones were identified based on their bullous microstructure, distinct from the smoother, uniform dermal bones forming the external skull. Using 3D World’s Smart Grid Tool, researchers defined regions of interest (ROIs) by adjusting grayscale intervals interactively, achieving precise segmentation of overlapping braincase fragments.

In tomographic sections through the eye socket and cheek region, these bullous tissues became clearly visible. 3D World’s segmentation and color-coded rendering allowed for visual differentiation between endocranial and dermal bones, providing a new 3D perspective of fossilized cranial anatomy.

Once segmented, each ROI was exported as a STL mesh for 3D reconstruction in Blender (4.5). This enabled the team to digitally reassemble scattered cranial elements and visualize the probable configuration of the original braincase.

Tomographic scan image of the Lepidotes skull showing a right parasagittal section through the cheek and eye socket regions. Presence of bony material is indicated in different greyscales. Endocranial bones differ from dermal bones by their bullous structure.
Tomographic scan of the Lepidotes skull highlighting endocranial bones in different colors.

Applications and Benefits

This research demonstrates the strong capabilities of Dragonfly 3D World in paleontological imaging and digital reconstruction. By combining µ-CT data with 3D World’s interactive visualization and segmentation tools, scientists can now analyze complex fossil morphologies non-destructively, preserving valuable specimens while extracting high-fidelity anatomical information.

The digital workflow—µ-CT acquisition, segmentation, and 3D modeling in Blender—serves as a replicable framework for studying other fossilized vertebrates. The approach offers measurable advantages:

  • Non-invasive visualization of hidden internal structures.
  • Accurate segmentation of superimposed or fractured bones.
  • 3D reconstruction of anatomically disarranged specimens for comparative morphology.

The study also provides evolutionary insights: the Lepidotes braincase, composed of separate ossifications rather than a unified bone case, shows similarities to its living relative, the bowfin (Amia), yet with greater extend of ossification and larger bones—suggesting evolutionary adaptation toward less robust endocranial framework.

Partially reconstructed braincase of Lepidotes from the Lower Jurassic of Germany. a - right lateral view, b - left lateral view, c - dorsal view, d - ventral view. Endocranial bony elements in a are prootic (left), sphenotic (middle) and pterosphenoid (right).

Summary

The Lepidotes braincase study exemplifies how modern imaging and visualization tools like Dragonfly 3D World are revolutionizing paleontological research. Through precise segmentation, flexible rendering, and export-ready data, 3D World bridges the gap between digital imaging and anatomical interpretation.

Researchers and institutions interested in using Dragonfly 3D World for paleontological or materials-based image processing can explore its capabilities through a FreeD license or contact the Dragonfly team to discuss applications in their own research.

Watch the segmentation process in action below, showing how Dragonfly 3D World enabled precise uncovering of the fossil braincase elements.

Acknowledgements

This study was made possible through collaborative efforts across research institutions. Micro-CT scanning was performed by Gabriel S. Ferreira, University of Tübingen, Germany. The research and findings were contributed by Detlev Thies and Jens Waschkewitz, Leibniz Universität Hannover, Germany.

Back to overview