Artificial intelligence-based prediction of boar reproductive fitness and health: Current status in research and practice

The development of artificial intelligence (AI) and machine learning (ML) is transforming reproductive management in boar studs by providing objective, data-driven methods of assessing and predicting sperm quality. Conventional sperm quality analysis is labor-intensive, sometimes subjective, and its ability to predict fertility outcomes is limited. Recent studies have shown that sperm quality parameters may be effectively analyzed using deep learning applied to imaging data, such as image-based flow cytometry and computer-assisted sperm analysis (CASA) data. Furthermore, ML methods based on Raman or near-infrared spectroscopy make it possible to non-invasively predict sperm viability during storage. Practical applications of AI in boar stud laboratories and under field settings include the prediction of future ejaculate quality through testicular ultrasound grey-scale analysis and systems for lameness detection and sperm transport monitoring. These applications expand the predictive scope to include boar reproductive fitness and health. While data standardization, model interpretability, and economic feasibility remain key challenges, the potential benefits and areas of application are broad. In this review, we summarize the current state of AI and ML innovations regarding boar reproductive fitness and health in research and practice. We highlight their contribution to the ongoing transition from descriptive sperm assessment to predictive, precise, and holistic reproductive management in the pig industry.