基于电子鼻检测的奶牛产后子宫炎早期预警模型初步研究

打开文本图片集

中图分类号：S857.23 文献标志码：A 文章编号：0366-6964（2026）01-0553-11

Abstract： The aim of this study was to evaluate the accuracy of the early warning model for metritis in dairy cows established by electronic nose detection. The detection of volatile organic compounds （VOCs） in feces and blood samples of 5O healthy dairy cows and 22 metritis cows before onset by electronic nose，preliminary analysis of data by orthogonal partial least squares discriminant analysis （OPLS-DA） model. The training set and test set were divided according to ，the early warning model was established and the performance of the electronic nose in predicting metritis was evaluated. The results showed that， the feces and blood of healthy and metritis dairy cows were clearly distinguished and clustered respectively. In this study feces are the best sample source for establishing an early warning model. Five different machine learning algorithms，namely decision tree （DT）， random forest（RF），K-nearest neighbors（KNN），eXtreme gradient boosting（XGBoost），and linear discriminant analysis （LDA），were used to build models. Predictive eficiency is analyzed in conjunction with model evaluation metrics and receiver operating characteristic curves （ROC）. It was found that the test set accuracy（ACC） reached O.88、O.96、0.88、0.91、O.98. Area under curve （AUC）was O.86、0.98、0.95、0.98 and 1.O0. As a result，the RF and LDA models perform the best in terms of predictive performance. The electronic nose has a high accuracy in predicting the occurrence of uterine infections.In conclusion，the detection of fecal VOCs of dairy cows before onset by electronic nose can achieve the purpose of early warning of metritis inflammation and has a great application prospect in the early diagnosis of dairy cow diseases.

Keywords： metritis; electronic nose; machine learning algorithm; early warning Corresponding author： HU Junjie，E-mail：hujj@gsau.edu.cn

子宫炎是奶牛产后高发的疾病之一，给全球奶牛养殖业造成了严重的经济损失。（剩余20379字）