ForamJ – A Tool for the Reproducible, Semi-automated Analysis of Foraminifera Micro Computed Tomography Datasets
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
The application of micro-computed tomography (µCT) to foraminiferal test analysis has opened new avenues for high-resolution, non-destructive 3D characterisation of internal and external morphological features. However, existing workflows are typically limited by low throughput, manual segmentation, and reliance on proprietary software, constraining dataset reproducibility and scalability. Here we present ForamJ, an open-source ImageJ plugin developed specifically for the semi-automated, reproducible analysis of foraminiferal µCT datasets. ForamJ streamlines key stages of image processing, including test segmentation, homogenous infill removal, chamber isolation, and morphological quantification, with outputs tailored to established micropaleontological metrics while providing a significant decrease in processing time versus existing methods. These include both calcite volume and chamber volumes, test surface area, inner and outer wall thickness, chamber centroid coordinates - parameters directly relevant to taxonomy, taphonomy, paleoecology, and biomineralisation processes. The plugin supports both single-sample and batch-processing workflows, enabling high-throughput analysis. Two use cases are presented: (1) the assessment of species-specific morphological variability in ancient cosmopolitan benthic foraminifera genus Cibicidoides spp. and (2) quantification of trends in between multiple species of modern planktonic foraminifera. Validation against manual segmentation of benthic foraminifera performed in commercial software ORS Dragonfly demonstrates ForamJ’s analytical fidelity, with a 1.30% difference in calcite volume between ForamJ and manually annotated tests, with a 0.9981% (± 0.001) segmentation accuracy and 0.97 (± 0.008) dice score with a median processing speed of 154 seconds per sample. By embedding foraminifera-specific metrics into an accessible, reproducible, and extensible framework, ForamJ provides a lightweight, dedicated digital toolset for advancing µCT-based test analysis in foraminiferal research.