(4) Beyond CASA: A New Framework for Sperm Motility Analysis

Abstract Authors: Kylie Cashwell¹; David Hart²; Calli Wilds¹; Ethan Carrow¹; Cameron A. Schmidt¹

1. Department of Biology, East Carolina University, Greenville, NC

2. Department of Computer Science, East Carolina University, Greenville, NC

Abstract Text: Male factor infertility accounts for 30–40% of infertility cases, with combined male and female factors adding another 20%. Because sperm function depends on reaching and fertilizing eggs, tracking motility is crucial for evaluating fertilization potential. IVF clinics commonly use Computer-aided Semen Analysis (CASA) for motility analysis, which typically relies on five-second microscopy snippet videos that often fail to capture extended, complex trajectories, path crossovers, and non-linear movements. This reliance on short tracking windows and abrupt velocity cutoffs leads to trajectory fragmentation and the loss of biologically relevant motility behaviors. To address these limitations, we developed a unified framework capable of tracking sperm for minutes at a time, performing advanced velocity analyses, resolving multi-path crossovers, and supporting more comprehensive measurements. We preserved the true sperm trajectories during crossover events by manually correcting mislabeled sperm identities, ensuring accurate CASA parameter outputs and reliable future measurements. This approach not only provides a better representation of standardized CASA parameters but also captures the complete sperm search mechanism. Subsequently, the unified framework will decompose head trajectories over extended durations allowing long-term motility measurements such as random walk properties and directional persistence lengths. We also intend to train a supervised machine learning model, such as a Recurrent Neural Network, using the manually corrected data to train complex tracking challenges like path crossovers and interactions with debris. By making the framework open-source and adaptable, researchers and clinicians can refine motility assessments, advance fertility diagnostics, and transcend current CASA limitations.