Study Reveals Microscopic Damage to Coral Skeletons from Disease

Researchers from Florida Atlantic University have pioneered a new method to analyze the microscopic structural damage caused by Stony Coral Tissue Loss Disease (SCTLD) in coral reefs. By combining X-ray microcomputed tomography (micro-CT) with deep learning algorithms, the team has produced detailed, three-dimensional insights into how the disease compromises coral skeletal integrity. The findings were published in the Journal of Structural Biology.

"Micro-CT gives us a window into the coral skeleton in a way that's never been possible before. By combining it with deep learning, we can automatically detect subtle changes in the skeleton caused by disease." — Alejandra Coronel-Zegarra, first author and Ph.D. candidate.

How the Research Was Conducted

The study focused on two key stony coral species: Montastraea cavernosa and Porites astreoides. Specimens included both healthy corals and those visibly affected by SCTLD.

The methodology relied on two advanced technologies:

• Micro-CT Imaging: This non-destructive technique was used to create high-resolution 3D reconstructions of coral skeletons. It revealed internal features like porosity, skeletal thickness, and structural orientation at a microscopic scale. Imaging was performed at the FAU High School Owls Imaging Lab.

• Deep Learning Analysis: To automate the analysis of complex images, researchers employed convolutional neural networks (CNNs). They specifically trained and tested three U-Net-based models—U-Net, U-Net++, and Attention U-Net—to automatically distinguish solid coral skeleton from pore spaces within the micro-CT scans.

Key Findings

The research produced significant results in two areas: the performance of the analytical method and the physical changes observed in the corals.

• All three deep learning models achieved accuracy rates exceeding 98% in differentiating skeleton from pores.
• The Attention U-Net model proved most efficient, completing a full image segmentation in seven hours, compared to 15 hours for the standard U-Net and 17 hours for U-Net++.

• The analysis revealed measurable differences in pore structure between healthy and diseased corals. These changes are believed to affect the overall skeletal integrity of corals with SCTLD.
• Structural differences were also noted between the two coral species, suggesting a potential link between a coral's inherent morphology and its vulnerability to disease at the microscopic level.

"Without high-resolution, 3D insights, scientists cannot fully understand how disease, warming oceans and other stressors compromise reef survival. Our analyses provide a clearer, quantitative picture of how environmental stressors reshape coral skeletons at the microscopic level." — Vivian Merk, Ph.D., corresponding author and assistant professor.

Merk also highlighted the broader potential of the methodology, noting it "opens new possibilities for analyzing other biological materials, engineered composites and even geological samples."

Background on the Disease and Research Impact

• Stony Coral Tissue Loss Disease (SCTLD) was first identified in Florida in 2014 and has since spread across the Florida Reef Tract and the Caribbean, causing significant mortality in reef-building corals.
• Traditional methods for studying microscopic coral features are often slow and can miss subtle changes. This new combined approach aims to address those limitations by providing faster, automated, and highly detailed analysis.

Funding & Recognition: The research was supported by the National Science Foundation and seed funding from FAU's College of Engineering and Computer Science and the FAU I-SENSE Institute. First author Alejandra Coronel-Zegarra received the 2025 Microscopy and Microanalysis Student Award for this work.

Study Co-Authors: Jamie Knaub (imaging lab assistant and Ph.D. candidate) and Abhijit Pandya, Ph.D. (professor in the Department of Electrical Engineering and Computer Science and Biomedical Science).