2024/10/23 by Andrea Nathansen, Matthis Clausen, Manuel Berning +6 · 1 voice · 1 citation
Physics and Astronomy · Biochemistry, Genetics and Molecular Biology · #Advanced X-ray Imaging Techniques #Digital Holography and Microscopy #Advanced Electron Microscopy Techniques and Applications
paper · pdf · doi:10.1117/12.3028309
openalex publication_date 2024/10/23 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/29
Biological soft tissues are functional agglomerates of cells. They constitute the microenvironment where intercellular communication occurs. In turn, their woven structure underlies mechanical properties that contribute to their roles in the context of the organs and the organisms that contain them. Therefore, determining the density and spatial distribution of cells within the tissue offers key information for understanding its physiological properties and its state. X-ray holographic nanotomography is a non-destructive imaging technique capable of resolving subcellular details in biological tissues that has shown promising advantages to study the structure of neuronal circuits. However, the dimensions of the datasets required – covering volume landscapes of ~mm<sup>3</sup> – make manual annotation of individual nuclei an unrealistic task. We developed and trained an automated image segmentation classifier that accurately detects and segments cell nuclei in mouse brain tissue imaged with x-ray holographic nanotomography, and that generalises to similar datasets obtained from biological replicates with minimal additional ground truth. It provides the spatial locations and morphologies of the ~80k nuclei per dataset with a high recall. It harnesses the strengths of a high-performance computing cluster and embeds the curated results in two main simplified outcomes: a data table and explorable image segmentations and meshes associated with the original dataset, in a browser-compatible format that simplifies proofreading by multiple users. The classifier we present here can be readily integrated into an automated analytical pipeline for histological datasets obtained with synchrotron x-ray holographic nanotomography in the context of systems neuroscience as well as broader tissue life science studies.