Framework Evaluates Predicted Sperm Trajectories In Crowded Videos

Researchers publish a framework to evaluate predicted sperm trajectories in crowded phase-contrast microscopy videos, with peer-reviewed article published Feb 10, 2026 in PLoS Computational Biology. The study adapts cell-tracking metrics, proposes sperm-specific modifications, releases a labeled dataset of 340 trajectories and code, and reports up to 30% improvement in tracking metrics, enabling more reliable long-term motility analysis.