D3492: Serial tomography and semi-automated analysis of insects

Investigations into the biology, evolution, and functional morphology of insects hold significant relevance not only for scientific inquiries but also for society at large. While genetic data collection has experienced substantial growth in recent years, the availability of comparative 3D morphological information has not kept pace. This discrepancy arises from the slow data acquisition of desktop-based systems, limited accessibility to synchrotron light sources, and the challenges associated with data analysis. Consequently, the majority of X-ray studies on insects rely on a very limited number of specimens.

To address these limitations, the imaging station at the KIT Light Source has been optimized for rapid X-ray imaging. Equipped with a high-speed detector system and a sample change robot, this setup enables serial micro-CT experiments and facilitates the scanning of large quantities of specimens. Recent experiments have demonstrated the method's efficacy in digitizing and performing 3D analyses of extensive scientific collections, encompassing both extant and fossil insects. Noteworthy findings include the examination of fossil parasites and the identification of unique mandibles in minute wasps.

Despite the progress made in X-ray imaging, the analysis of insect micro-CT data remains arduous and time-consuming. Manual segmentation of voluminous and intricate datasets, in particular, continues to present a significant bottleneck, especially for large-scale comparative studies. The application of Biomedisa's semi-automatic segmentation can significantly reduce the manual workload involved in analyzing extensive and complex datasets. Moreover, it has the potential to train data for neural networks, enabling fully automated segmentation in the future.