Journal of Poultry Sciences and Avian Diseases

Artificial intelligence (AI), machine learning, computer vision, sensor-based systems, and Internet of Things (IoT) technologies are increasingly used for automated monitoring of poultry health and welfare, but reported model performance varies across tasks, data sources, validation strategies, and production settings. This systematic review and meta-analysis evaluated the reported performance of AI-based models for poultry health and welfare monitoring. PubMed/MEDLINE, Scopus, and Web of Science were searched for peer-reviewed studies involving poultry, reporting health- or welfare-related monitoring outcomes, and providing extractable model-performance data. Data were extracted independently by two reviewers, and pooled accuracy was estimated using a Restricted Maximum Likelihood random-effects model in STATA version 17. Twenty-one studies published between 2012 and 2025 were included, of which ten were eligible for meta-analysis. Included systems used convolutional neural networks, random forests, support vector machines, and other machine-learning approaches for behavior recognition, disease detection, lesion assessment, mortality detection, environmental monitoring, and welfare assessment. The pooled accuracy estimate was 90.39% (95% confidence interval: 84.89–95.89; P < 0.001), with substantial heterogeneity among studies. Reported performance varied by application domain, data source, model type, assessment dimension, sample size, and integration with IoT-based systems. AI-based technologies show promise for automated poultry health and welfare monitoring; however, heterogeneous methods and limited external validation restrict the generalizability of pooled estimates. Future studies should prioritize standardized reporting, open datasets, external validation, and testing under commercial farm conditions.

H5N1, or highly pathogenic avian influenza, continues to pose a serious risk to public health and poultry. In order to differentiate H5N1-infected from healthy chicken samples and to guide the development of diagnostics and vaccines, we postulated that Machine Learning (ML) applied to RNA sequencing (RNA-Seq) data could detect biologically meaningful immune gene signatures. While ML has been applied to other avian RNA-Seq datasets, to our knowledge, this study represents one of the first applications of interpretable ML to chicken H5N1 RNA-Seq transcriptomic data for immune signature discovery.

RNA-Seq data from H5N1-infected chicken lung and ileum tissues (ArrayExpress E-MTAB-2908) that were publicly available, were chosen. Fivefold stratified cross-validation (~80/20 train/test per fold) was used to train two supervised ML models, such as Random Forest (RF) and linear-kernel Support Vector Machine (SVM). Performance was evaluated using Area Under Curve (AUC) and Receiver Operating Characteristic (ROC) curves.

RF reached a mean AUC = 0.85, while SVM-Linear achieved AUC = 0.75. Top-ranking interferon-stimulated genes (ISGs), including IFIT5, MX1, and OASL, were consistently upregulated in infected samples, indicating activation of type I interferon pathways. Concordant findings across models support the stability and biological relevance of the identified signatures despite the modest sample size.

These genes are highlighted as potential candidates for early molecular diagnostics and for tracking vaccine-induced antiviral immunity. These findings broaden the methodological applications of interpretable ML in avian transcriptomics and computational immunology.

This study aimed to evaluate the physiological relationship between body weight, selected biochemical and hormonal serum parameters, and their association with productive performance in ISA Brown laying hens during the mid-laying period (42–50 weeks of age). A total of 135 ISA Brown laying hens were into three experimental groups according to body weight category: light, medium, and heavy body weight, with 45 hens per treatment. Productive performance traits, including hen-day egg production, egg weight, and egg mass, were recorded. In addition, serum concentrations of reproductive hormones (FSH, LH, and estradiol) and selected biochemical parameters, including glucose, total protein, albumin, cholesterol, and triglycerides, were determined.

The results demonstrated that medium-weight hens showed significantly higher (P ≤ 0.05) hen-day egg production and increased concentrations of reproductive hormones, total protein, and albumin compared with the other groups, indicating a more favorable physiological and metabolic balance associated with improved reproductive efficiency and laying persistency. In contrast, heavy-weight hens recorded significantly higher (P ≤ 0.05) egg weight, glucose, cholesterol, and triglyceride concentrations, reflecting increased metabolic activity and fat deposition associated with greater body weight. Pearson correlation analysis within body weight categories revealed significant positive correlations (P ≤ 0.05) between reproductive hormones and productive performance traits, particularly in medium-weight hens, whereas elevated cholesterol and triglyceride levels were negatively associated with some productive parameters.

In conclusion, maintaining moderate body weight in ISA Brown laying hens contributes to improved physiological balance and enhanced productive performance during the mid-laying period.

This paper develops an automated non-invasive system for disease detection in poultry farming by classifying chicken droppings. This study examines multiple deep learning architectures employing transfer learning to classify chicken droppings. Based on comparative analysis, a knowledge distillation network was proposed utilizing DenseNet as the teacher model due to its superior performance metrics. A lightweight student network was proposed and trained using knowledge distillation. This effectively transfers discriminative feature representations from the teacher model while significantly reducing computational complexity. Experimental evaluation on a publicly available poultry faecal image dataset demonstrated that the proposed knowledge distillation network achieved an accuracy of 98.40%, precision of 97.72%, specificity of 99.44%, recall of 98.21%, and F1-score of 97.96%.

This study investigated the effects of graded dietary digestible methionine plus cystine (dMC) on hepatic and ovarian gene expression, reproductive hormones, and antioxidant status in post-molt Hy-Line W-80 laying hens (79 to 95 weeks of age; n = 180). Hens were randomly allocated to five isonitrogenous and isoenergetic diets providing dMC: Lys ratios of 75, 83, 91, 99, and 107 (dMC75 to dMC107), corresponding to dMC levels of 0.490% to 0.698%. Hepatic mRNA expression of yolk precursor genes (VTG1, VTG3, VLDLR), apolipoproteins (APOA1, APOA4), and lipogenic enzymes (DGAT2, LPIN1) showed quadratic responses, with maximal expression at dMC99. Ovarian mRNA expression of growth factors (IGF1, IGF1R, BMP6), folliculogenesis markers (AMH, INHA, INHBA), angiogenic factors (VEGFA, KDR), and transcription factors (GATA4, FOXL2) followed similar quadratic patterns, peaking at dMC99. Circulating estradiol, progesterone, insulin-like growth factor-1, and anti-Müllerian hormone exhibited significant quadratic responses, with peak values generally observed at dMC99. Malondialdehyde decreased from 6.49 to 3.55 nmol/mL, while total antioxidant capacity increased from 1.11 to 1.77 mmol/L, and catalase activity rose from 44.62 to 78.32 U/mL. Principal component analysis separated treatments along PC1 (54.9% variance), with dMC99 clustering with favorable parameters. Strong positive correlations were observed among estradiol-17β, catalase, anti-Müllerian hormone, and lipid metabolites (r = 0.78 to 0.92). These results suggest that dietary sulfur amino acid supply is associated with quadratic regulation of the hepatic-ovarian axis in post-molt laying hens, with optimal effects observed at approximately 0.648% dMC (dMC: Lys = 99).

This study compared the preliminary serum immune-marker responses induced by two vaccine preparations against avian pathogenic Escherichia coli (APEC) in broiler chickens: a bacterial ghost preparation (BGs) and a liposome-based preparation (LIPO). Serum IL-4, IL-10, IL-12, IL-1β, IFN-γ, and total IgY were measured by ELISA to evaluate humoral, cellular, and inflammatory responses relative to a negative control group. Overall differences among the three groups were statistically significant for all measured markers (p < 0.0001). Direct comparison of the vaccinated groups showed significantly higher IL-10, IL-1β, and total IgY concentrations in the BGs group, whereas IL-4 was significantly higher in the LIPO group. IL-12 and IFN-γ did not differ significantly between the two vaccinated groups. These findings indicate that the two preparations generated distinct preliminary serum immune-marker profiles. Because protective efficacy was not evaluated in an APEC challenge model, further studies are required before either preparation can be considered an effective vaccine against colibacillosis.

Green banana powder is a functional feed ingredient rich in resistant starch and dietary fiber, with potential applications in poultry nutrition. However, information regarding its use in indigenous Vietnamese chicken breeds under practical tropical rearing conditions remains limited. This study evaluated the effects of dietary green banana powder supplementation on growth performance, feed efficiency, watery droppings incidence, and litter quality in H’Mong chickens from 1 to 56 days of age. A total of 120 one-day-old H’Mong chickens were randomly assigned to four dietary treatments containing 0%, 1%, 2%, or 3% green banana powder, with three replicates of 10 chickens per treatment. Dietary supplementation significantly affected final body weight, average daily gain, feed conversion ratio, watery droppings incidence, wet-dropping days, litter moisture, litter quality score, and odor score (P < 0.05). Chickens receiving 2% green banana powder showed the best productive and litter-related responses. Feed intake and survival rate were not significantly influenced by dietary treatment (P > 0.05). Under the conditions of the present study, supplementation with 2% green banana powder improved feed efficiency, productive performance, watery droppings-related indicators, and litter condition in H’Mong chickens raised under practical tropical farm conditions.

.