Artificial intelligence in Andrological flow cytometry: The next step?

• Published in: Animal reproduction science 2024-11, Vol.270, p.107619, Article 107619
• Main Authors: Peña, Fernando J., Martín-Cano, Francisco Eduardo, Becerro-Rey, Laura, da Silva-Álvarez, Eva, Gaitskell-Phillips, Gemma, Ortega-Ferrusola, Cristina, Gil, María Cruz
• Format: Article
• Language: English
• Subjects: Andrology - methods; Animals; Artificial Intelligence; Equine; Flow cytometry; Flow Cytometry - methods; Male; Self-learning; Spermatozoa; Spermatozoa - physiology; Stallion
• ISSN: 0378-4320; 1873-2232; 1873-2232
• DOI: 10.1016/j.anireprosci.2024.107619

Since its introduction in animal andrology, flow cytometry (FC) has dramatically evolved. Nowadays, many compartments and functions of the spermatozoa can be analyzed in thousands of spermatozoa, including, but not limited to DNA, acrosome, membrane integrity, membrane symmetry, permeability, and polarity; mitochondrial mass and mitochondrial membrane potential, identification of reactive oxygen species, ion dynamics, and cellular signaling among many others. Improved machines, many more probes, and new software are greatly expanding the amount of information that can be obtained from each flow cytometry analysis. Modern flow cytometers permit the simultaneous investigation of many different sperm compartments and functions and their interactions, allowing the identification of sperm phenotypes, helping to disclose different sperm populations within the ejaculate. Complex flow cytometry panels require a careful design of the experiment, including selecting probes (fully understanding the characteristics and properties of them) and adequate controls (technical and biological). Ideally, compensation and management of data (“cleaning”, transformations, the establishment of gates) are better performed post-acquisition using specific software. Data can be expressed as a percentage of positive cells (typically viability assays), intensity of fluorescence (arbitrary fluorescence units, i.e. changes in intracellular Ca2+) or dim and bright populations (typically assays of membrane permeability or antigen expression). Furthermore, artificial intelligence/self-learning algorithms are improving visualization and management of data generated by modern flow cytometers. In this paper, recent developments in flow cytometry for animal andrology will be briefly reviewed; moreover, a small flow cytometry experiment will be used to illustrate how these techniques can improve data analysis. •All aspects of flow cytometry have evolved in the last decade.•Modern flow cytometers can provide a huge amount of data from each single analysis.•Management of single-cell data provided benefits from the use of artificial intelligence/self-learning algorithms.