A deep learning-enabled toolkit for the 3D segmentation of ventricular cardiomyocytes

Clinical Snapshot

Key Findings

• P Value: Not applicable — performance benchmarking study; no hypothesis testing reported

• Effect Size: Adapted Rand error of 0.063 ± 0.034, corresponding to approximately 94% voxel-pair agreement between automated segmentation and expert annotation

• Primary Outcome: Automatic 3D cardiomyocyte segmentation accuracy on held-out test set, measured by adapted Rand error

• Nnt Or Sensitivity: Semi-automatic workflow throughput: 3 cells per minute on a challenging, previously unseen dataset; no sensitivity/specificity metrics reported in the abstract

• Confidence Interval: Not formally reported; variability expressed as standard deviation (±0.034)

Clinical Application

Open-source availability with a dedicated graphical user interface makes the toolkit accessible to laboratories without dedicated computational expertise. Semi-automatic throughput of 3 cells per minute is operationally feasible for research-scale analyses. Deep learning inference requires GPU-capable hardware, which is increasingly standard in research institutions but may require infrastructure investment in smaller laboratories. This toolkit has direct relevance to Australian cardiac research institutions, including those affiliated with the National Heart Foundation of Australia and university-based cardiac physiology groups. It is not a clinical diagnostic tool and therefore does not require TGA approval or PBS listing. It is relevant to NHMRC-funded preclinical cardiac research and may support Australian contributions to international cardiac cell modelling consortia. The inclusion of human cardiac tissue data is particularly relevant to Australian biobanking and cardiac surgical research programs. RACGP guidelines are not directly applicable, as this is a research tool rather than a clinical intervention. Preclinical and translational cardiac researchers working with ventricular myocardial tissue from any of the seven validated species; particularly relevant to groups studying cardiac remodelling in myocardial infarction, heart failure, and cardiomyopathies

Abstract

Segmentation of cardiomyocytes in microscopic 3D volumes is key to our understanding of cardiac (patho-)physiology; however, it poses substantial experimental and analytical challenges. Therefore, researchers often resort to inferring 3D information from 2D segmentations, which can lead to biased and incorrect conclusions. Deep learning-based methods are showing promise with respect to robustly segmenting objects in volumes acquired using various imaging modalities; yet, they have not been applied to high-resolution 3D cardiomyocyte segmentations, and suitable open-source tools and datasets are lacking. Here, we present a deep learning-enabled toolkit for segmentation of individual cardiomyocytes in 3D confocal microscopy volumes. We include a dataset of 73 volumes with expert annotations, covering seven species, including mouse, human, and elephant, and containing samples generated under different experimental conditions, such as post-myocardial infarction and ex vivo slice cultures. The toolkit additionally contains an image restoration workflow to address imaging-related artefacts, such as spatially varying blur. Our automatic cardiomyocyte segmentation workflow achieved an adapted Rand error of 0.063 ± 0.034 (∼94% voxel-pair agreement) on the test set. Our semi-automatic workflow reached a throughput of 3 cells min-1 on a challenging, previously unseen dataset. The toolkit and data are open-source and accessible through a dedicated graphical user interface. In summary, we provide an accessible toolkit enabling researchers to extract quantitative data on cardiomyocyte microstructure from 3D confocal image stacks of cardiac tissue. Given the size and diversity of our dataset, we expect our methods to perform well across species and experimental conditions, facilitating high-quality 3D reconstructions of large numbers of individual cardiomyocytes. KEY POINTS: 3D cardiomyocyte microstructure is a key determinant of cardiac function in health and disease. However, reliable extraction and quantification of 3D cardiomyocyte cytoarchitecture pose significant experimental and computational challenges. We present an effective experimental protocol and a deep learning-enabled toolkit for sample preparation and 3D analysis of cardiomyocyte morphology in ventricular myocardium. Our method is validated across seven species (mouse to human) and in samples prepared in diverse experimental conditions from a range of models, including myocardial infarction and ex vivo tissue culture, highlighting the robustness and versatility of our workflow. Our open-source dataset and toolkit enable large-scale analyses and extraction of realistic 3D geometries of ventricular microstructure. These can be used to explore a host of research questions and provide a new resource for modelling cardiac function at the cellular level.

References

1. 1.Greiner, J., Frangiamore, F., Sonak, F., Madl, J., Seidel, T., Kohl, P., & Rog-Zielinska, E. A. (2026). A deep learning-enabled toolkit for the 3D segmentation of ventricular cardiomyocytes. The Journal of Physiology. https://doi.org/10.48550/arXiv.1804.03999