Digital technologies that continuously quantify animal behavior, physiology, and production offer significant potential for the early identification of health and welfare disorders of dairy cows. In this study, a multimodal artificial intelligence (AI) framework is proposed for real-time health monitoring of dairy cows through the integration of physiological, behavioral, production, and thermal imaging data, targeting veterinarian-confirmed udder, leg, and hoof infections. Predictions are generated at the cow-day level by aggregating multimodal measurements collected during daily milking events. The dataset comprised 88 lactating cows, including veterinarian-confirmed udder, leg, and hoof infections grouped under a single ‘sick’ label. To prevent information leakage, model evaluation was performed using a cow-level data split, ensuring that data from the same animal did not appear in both training and testing sets. The system is designed to detect early deviations from normal health trajectories prior to the appearance of overt clinical symptoms. All measurements, with the exception of the intra-ruminal bolus sensor, were obtained non-invasively within a commercial dairy farm equipped with automated milking and monitoring infrastructure. A key novelty of this work is the simultaneous integration of data from three independent sources: an automated milking system, a thermal imaging camera, and an intra-ruminal bolus sensor. A hybrid deep learning architecture is introduced that combines the core components of established models, including U-Net, O-Net, and ResNet, to exploit their complementary strengths for the analysis of dairy cow health states. The proposed multimodal approach achieved an overall accuracy of 91.62% and an AUC of 0.94 and improved classification performance by up to 3% compared with single-modality models, demonstrating enhanced robustness and sensitivity to early-stage disease.

This study evaluated infrared thermography (IRT) as a dual-purpose, non-invasive tool for detecting limb pathology and subclinical mastitis in dairy cows. Surface temperatures were measured at the coffin, pastern, and fetlock regions of all limbs and at the teat surface of all udder quarters. Limb health status was determined by orthopedic clinical examination, and mastitis status was assessed using the California Mastitis Test (CMT). Hindlimbs were significantly warmer than forelimbs in healthy cows (p < 0.001), with confirmed bilateral symmetry (r = 0.524–0.569). Limbs with pathology showed significantly higher temperatures at all anatomical sites, with particularly large effect sizes in the hindlimbs (r around 0.80). Additional lesion-specific analyses demonstrated that both infectious and claw horn lesions exhibited significantly higher temperatures than healthy limbs. In the hindlimbs, infectious lesions showed higher thermographic values than claw horn lesions. CMT-positive quarters showed teat temperatures of 30.0–30.2 °C compared with 25.2–25.8 °C in CMT-negative quarters (p < 0.001). Receiver operating characteristic (ROC) analysis demonstrated excellent diagnostic performance for teat surface temperature in identifying CMT-positive quarters, with an area under the ROC curve (AUC) of 0.956. The optimal threshold was 29.5 °C, corresponding to a sensitivity of 0.992 and a specificity of 0.838. Each 1 °C increase in teat temperature increased the odds of CMT positivity by 5.3-fold (p < 0.001). A practical three-zone thermographic framework was established: <27 °C (healthy), 27–29.5 °C (monitoring recommended), and ≥29.5 °C (mastitis likely). Teat temperature was not associated with composite milk somatic cell count, but hindlimb pathology was independently associated with elevated teat temperatures (partial r = 0.23–0.26, p ≤ 0.001). These findings demonstrate that IRT has strong potential as an integrated screening tool for dairy cow limb health and udder health monitoring.

Rumen microbial fermentation plays a central role in nutrient utilization and milk production in dairy cows. This study evaluated the effects of supplementation with exogenous fibrolytic enzymes, alone or in combination with live yeast on rumen microbiota, fermentation characteristics, nitrogen-related metabolites, and production performance during the transition from outdoor grazing to indoor housing. Thirty Lithuanian Red dairy cows were assigned to control (CTR), enzyme (E), or enzyme plus yeast (YE) treatments across outdoor (OD) and transit (T) periods, while nine cows (three per group) were selected for rumen and microbiota analysis. Rumen bacterial communities were characterized using 16S rRNA gene sequencing, and functional parameters were evaluated using linear mixed-effects models. Supplementation resulted in selective changes in several bacterial genera, including Blautia spp., WPS-2, Ruminococcus spp., Erysipelotrichaceae UCG-009, Sharpea spp., uncultured Bacteroidales, and Prevotellaceae UCG-003, and was associated with alterations in fermentation patterns, particularly propionate concentration, and in nitrogen metabolism, including putrescine dynamics. The transition period significantly influenced microbial diversity and total bacterial abundance across treatments. Cows in the YE group maintained higher milk yield during the transition period. Overall, dietary supplementation modulated specific rumen metabolic responses and contributed to production stability without causing large-scale changes in overall microbial structure.