Rapid analysis of chinese blanched chicken (Mahuang, Tuer, and Huangyou) based on volatile compounds and machine...

To promote the intelligent development of the poultry industry, this study systematically analyzed volatile organic compounds (VOCs) in three Chinese Blanched Chicken (CBC) breeds (Mahuang, Tuer, and Huangyou) using gas chromatography-ion mobility spectrometry (GC-IMS). A machine learning-based CBC classification model was developed for rapid, precise breed identification. Based on GC-IMS, 66 VOCs were identified, with 54 key compounds selected through Extra-Trees and F-regression algorithms. Ethyl acetate and 3-hexen-1-ol were the marker VOCs. The neural network demonstrated exceptional performance (prediction accuracy = 0.9611), significantly outperforming XGBoost, decision tree, and random forest models. The results indicated that VOC-based rapid differentiation technology enabled efficient CBC breed identification, and the NN model offered an innovative solution for precise classification. This research provides robust technical support for intelligent upgrading in China's poultry processing sector, enhancing quality control and production efficiency through data-driven methodologies.