3D+t human sperm flagellum centerline dataset.

Area of Science:

• Reproductive Biology
• Biomedical Imaging
• Computational Biology

Background:

• Human sperm motility is vital for fertility, but traditional 2D analyses miss complex 3D flagellar movements.
• Limited availability of comprehensive 3D+time sperm motility datasets hinders research progress.

Purpose of the Study:

• Introduce 3D-SpermFlagella, the first large-scale 3D+time dataset of human sperm flagellum centerline annotations.
• Provide a resource for studying sperm motility in full spatial complexity and for AI model development.

Main Methods:

• Utilized multifocal video microscopy to capture sperm movement.
• Annotated flagellar centerlines in 3D+time for 135 human sperm samples.
• Included data under non-capacitating (NCC) and capacitating (CC) conditions, with coordinates in micrometers and voxels.

Main Results:

• Established the 3D-SpermFlagella dataset, featuring detailed 3D+time flagellar centerline data.
• Dataset includes annotations for sperm under both NCC and CC conditions.
• Provides spatio-temporal data crucial for advanced motility analysis.

Conclusions:

• The 3D-SpermFlagella dataset is a significant resource for advancing the study of human sperm motility.
• Facilitates the development and benchmarking of AI-driven sperm tracking and segmentation models.
• Enables deeper insights into the biomechanics of sperm movement and its role in fertility.