Application of artificial intelligence and machine learning in bovine respiratory disease prevention, diagnosis, and classification

Bovine respiratory disease (BRD) is the leading infectious disease in cattle, resulting in significant economic losses and welfare concerns in beef and dairy production systems. Traditional diagnostic methods for BRD typically rely on clinical observations and diagnostic laboratory tests, which can be time consuming with moderate diagnostic sensitivity. In recent years, machine learning (ML) and AI have emerged as powerful tools in animal health research, offering opportunities for improving BRD diagnostics and management. This review explores the current landscape of published literature on the use of ML and AI in BRD prevention, diagnostics, and classification. First, disease classification and pathogen identification models leveraging supervised models and metagenomic sequencing have identified specific community structure information in classifying specific BRD cases. From epidemiological datasets tracking disease outbreaks and risk factors, user-friendly platforms for producers and veterinarians are capable of being generated and deployed, providing customized scenarios, potential economic impacts, and pathogenic effects as a decision-support tool. Veterinarian-operated technologies, such as computer-aided lung auscultation stethoscopes, can automatically calculate lung scores and associated BRD severity likelihoods. Prediction and detection models used to leverage physical characteristics and feed consumption data provide novel methods of categorizing BRD risk. Finally, sensor technology monitoring behavioral or motion-based information provides continuous data on animal health and can enable early automated detection of BRD symptoms. Through synthesizing research in these key areas, this narrative review highlights the transformative potential of AI and ML in improving the accuracy, speed, and efficiency of BRD diagnostics, enhancing disease control and cattle welfare.