农林科学最新文献

Veterinary forensic science is a specialized discipline that investigates non-natural animal deaths and injuries. Determining the cause of death in free-roaming cats often presents challenges due to underdeveloped pattern recognition methods and the intricate interplay of events and contributing factors. The integration of natural language processing into forensic investigations may offer potential for the computerized automation of document analysis, with the goal of supporting investigational decision-making. The objective of this study was to develop an AI framework leveraging a transformer-based large language model (LLM) to classify veterinary forensic autopsy reports derived from 271 feline autopsies and to identify the causes of death. We first identified discriminative lexical patterns between vehicular trauma (VT; n = 111) and non-VT cases (n = 160) using Term Frequency-Inverse Document Frequency analysis. The best-performing model, based on Clinical-Longformer, achieved an accuracy of 0.945, precision of 0.885, recall of 1.000, F1 score of 0.939, and AUC-ROC of 0.950. Model interpretability using Integrated Gradients revealed that VT cases were characterized by terms associated with traumatic physical injuries, whereas non-VT cases exhibited a broader range of diagnostic descriptions. These findings demonstrate the feasibility of applying LLMs to veterinary forensic pathological texts and underscore the importance of domain-specific token interpretation for model explainability.

Prohibitive costs associated with high-field MRI systems have resulted in reduced access in resource-limited areas with consequential healthcare inequities. This has renewed interest in purposefully designed low-field (LF) systems that, despite inherently lower signal-to-noise ratios, have advantages in terms of cost, portability, and accessibility. This pilot study successfully translated a previously developed LF (0.05 T) MRI canine cadaver brain protocol for in vivo application and compared the acquired images with paired 1.5 T images from 21 different canine patients. The visibility of 15 anatomic features was ordinarily scored (scale 1-3) on transverse T1-weighted post-contrast (21/21 patients) and T2-weighted (17/21 patients) sequences. Lateral ventricles were easily visualized (89.3%-100% scored 3/3) across all sequence‒system combinations, with T2-weighted sequences also providing good identification of the mesencephalic aqueduct (56.3%-100%), fourth ventricle (77.1%-100%), and thalamus (68.8%-100%). Absolute visual grading characteristics (VGC) analysis confirmed comparable performance of the LF system to the 1.5 T MRI for these features, a particularly relevant finding given the system's originally intended application for pediatric hydrocephalus neuroimaging in resource-limited areas. This study represents the first in vivo usage and evaluation of a 0.05 T MRI in a clinical veterinary context.

PAMK are heterogeneous glycans with immunomodulatory potential, but their structural basis and mechanisms in regulating thymic injury remain unclear. Structural characterization revealed that PAMK is mainly composed of arabinose, galactose, rhamnose, galacturonic acid, and glucose, featuring a dominant backbone consisting of →4)-α-D-GalpA-(1→, →4)-β-D-Galp-(1→, →2)-α-L-Rhap-(1→, and →3,4)-α-D-GalpA-(1→ with highly branched side chains. This complex structure provides the biochemical foundation for its bioactivity. To explore its immunoprotective effects, we established a CTX-induced thymic necroptosis model in goslings. CTX markedly disrupted cytokine profiles and activated necroptosis signaling, accompanied by downregulation of novel-miR-2 and upregulation of TRADD and MLKL. Dietary PAMK supplementation significantly restored cytokine homeostasis and inhibited necroptosis pathway activation. Mechanistically, PAMK upregulated novel-miR-2, which negatively regulated TRADD and MLKL expression, thereby attenuating necroptosis signaling. Overexpression of novel-miR-2 suppressed TRADD/MLKL-mediated necroptosis, whereas its inhibition produced the opposite effect. Furthermore, PAMK mitigated oxidative stress and necroptotic cell death induced by TRADD and MLKL overexpression. Collectively, our findings demonstrate that the defined structural features of PAMK underpin its ability to protect against CTX-induced thymic necroptosis by upregulating novel-miR-2 and suppressing the TRADD/MLKL axis. This work links PAMK's composition to its immunoprotective function and highlights its potential as a safe immunoregulatory agent in poultry.

While dogs' behavioural indicators of psychological stress are widely used, their function as reliable stress markers remains debated. Using the Strange Situation Procedure (SSP) as a social-stress paradigm, we evaluated contextual correlates of dog behaviours to explore their validity as stress indicators. In 64 dogs, behaviours commonly used as stress markers were video-coded and analysed via GLMMs to investigate differences across test episodes and attachment styles (secure, anxious, avoidant), while salivary cortisol, oxytocin, and rectal temperature provided a physiological framework. Episode had significant effects on shaking (p < 0.05), lip-licking (p < 0.001), paw-lifting (p < 0.001), whining (p < 0.001), head-turning (p < 0.001), sniffing the door (p < 0.001), and locomotion (p < 0.001), while attachment style significantly affected head-turning only (p = 0.004). Our findings suggest that the observed behaviours were not expressed linearly in relation to expected stress levels across the SSP episodes. This lack of linearity, combined with the divergent physiological correlates found across attachment groups-for instance, oxytocin correlated positively with lip-licking in securely attached dogs (p < 0.05), but negatively in anxious ones (p < 0.05) -, suggests that these behaviours do not function as simple stress indicators. Rather, their meaning appears to be contingent upon the individual's internal state and the socio-environmental context. We discuss our results in light of an alternative semiotic interpretative framework.

Detailed analysis of stabled horse behaviour can reveal accurate information about its well-being. Advances in deep learning now allow these behaviours to be tracked without being invasive through the use of video data. This study evaluated a convolutional neural network for recognising standing, lying, and drinking behaviours in a horse housed in a wooden stall and recorded continuously over 29 consecutive days. Model predictions were compared with manually annotated ground truth data. Standing was detected with high precision (97.5%) and high recall (89.2%). Lying behaviour was classified with high precision (92.8%) but lower recall (63.1%). Activity patterns showed that standing dominated daily time budgets (>85%), lying accounted for 5-10%, and drinking occurred most often between 04:00 pm and 10:00 pm. These results demonstrate that deep learning can classify common equine behaviours from video, supporting its use in automated welfare monitoring. Future evaluations will explore the recognition of less frequent behaviours.

The objective was to evaluate the potential of machine learning algorithms utilizing thermal imaging and/or milk-based parameters for the identification of subclinical mastitis(SCM) in dairy cows. Holstein dairy cows (n = 194) were enrolled, and milk samples were collected from each quarter separately (n = 776 quarter-level samples) and analyzed for milk-based parameters. Four images per cow, representing all four quarters, were captured using a handheld infrared camera. Skin temperatures were extracted using Fluke Connect software, yielding 776 images for analysis. SCM was defined at the quarter level as SCC > 200,000 cells/mL, identifying 139 infected and 637 healthy quarters. The data were analyzed to test Random Forest Classifier(RFC), Logistic Regression Classifier, AdaBoost Classifier(ABC), and Naïve Bayes Classifier algorithms. The RFC models using milk features and temperature features achieved accuracy and precision scores of 0.88 and 0.88, respectively, yielding 0.90 AUC values, with high feature importance for milk lactose and SNF. Using only thermal imaging features, the ABC model yielded accuracy and precision of 0.82 and 0.67, respectively, with 0.56 AUC. We conclude that SCM prediction using machine learning algorithms is most promising when combined with thermal images and milk-based parameters. Future studies with larger datasets and refined methods for thermal image capture and analysis are warranted to validate these findings.

The transition from gestation to lactation is a critical period that leads to insulin resistance (IR) in dairy cows. IR detection requires invasive sampling and laborious methods; therefore, novel and non-invasive approaches are essential. Extracellular vesicles (EVs) are nano-sized particles released by cells and are linked to disease pathophysiology. Hence, the present study aimed to explore the potential application of cow milk EVs (MEVs) in detecting IR in dairy cows. An intravenous glucose tolerance test (IVGTT) was performed (n = 12), and the cows were classified as low, moderate, or high IR based on prepartum area under the curve insulin values. MEVs were enriched using double-size exclusion chromatography and physico-chemically characterized. The results revealed that the particle diameter of MEVs was significantly higher (p ≤ 0.05) in the high-IR group compared to other groups. Out of 26 fatty acids (FAs) detected in MEVs, three long-chain FAs differed significantly in abundance (p ≤ 0.05) in high-IR group. Proteomic analysis revealed a correlation of inflammation and the IR-related proteins; Annexin A1 and Rab proteins, to IR severity. Overall, this study identified promising predictors of IR in cows, supporting the development of non-invasive biomarkers for early detection, facilitating improved herd and disease management, and reducing economic losses.

Significant causes of morbidity in veterinary medicine are tumourigenic diseases and pathogenic infections, which pose significant threats to animal health. Traditional methods for treating tumours and bacterial infections primarily involve chemotherapy and antibiotic therapy, both of which are associated with drug resistance and adverse effects. Photodynamic therapy (PDT) is gaining increasing interest in veterinary clinical practice because of its distinctive benefits such as spatiotemporal selectivity, low toxicity, fewer negative effects, and reduced risk of drug resistance. This review summarises the fundamental principles of PDT and its applications in combating tumours and microbial infections in veterinary practice involving companion animals, livestock, wildlife, and reptiles. It is worth noting that the existing clinical evidence is limited by very small sample sizes and the absence of comparator groups; however, these studies have provided a preliminary theoretical foundation and technical reference for further clinical development of PDT. Moreover, the opportunities and challenges associated with PDT in veterinary clinical practice are discussed herein.

In this study, Salmonella Typhimurium (ST) vaccine diluent effects on caecal tonsils gene regulation in layer chicks were investigated. A live attenuated ST vaccine was reconstituted in Marek's and/or water diluents, chicks vaccinated and caecal tonsils were assessed for global gene expression at days 7 and 14 post-vaccination. Differentially expressed genes (DEGs) involved in the activation of the intestinal immune network for IgA production were upregulated earlier (e.g. day 7 post-vaccination) in the water diluent group, while in the Marek's diluent group, this pathway was enriched by the upregulated DEGs at day 14 post-vaccination. The commonly upregulated biological pathways between the two diluents were cytokine-cytokine receptor interaction, toll-like receptor signaling pathway and cytosolic DNA-sensing pathway. At day 7 post-vaccination, 28.5% and 40.2% of the significantly upregulated genes were unique to Marek's and water diluents, respectively, whereas 31.3% were common between the Marek's and water diluent groups, showing the role of the diluent in priming the vaccine for host immune system modulation. At day 14 post-vaccination, 52.7% and 15.4% of the upregulated DEGs were unique to Marek's and water diluents, respectively. Overall, the data showed that the live attenuated ST vaccine reconstituted in Marek's diluent affected the regulation of immune system-related genes in the caecal tonsils of chicks, with a difference in pattern when water was used as a diluent.