ANN-GA optimized ultrasound-microwave synergistic extraction of Procambarus clarkii head oil: comprehensive lipidomics and volatile organic compounds analysis.

This study aimed to efficiently and environmentally extract oil from Procambarus clarkii heads using ultrasound-microwave synergistic assisted n-hexane (UMNH) and optimize the extraction process through response surface methodology (RSM) and neural network model genetic algorithm (ANN-GA). Compared with the n-hexane (NH) method 39.01 ± 1.61% oil recovery, the optimized UMNH method reached 97.10% oil recovery under 475 W ultrasonic power, 830 W microwave power, 1:16 g/mL solid-liquid ratio, and 25 min extraction time. Among 1964 lipid species, glycerides accounted for 82.98%-90.88% of total lipids, with increased triglyceride (TG)-type eicosapentaenoic acid and docosahexaenoic acid in the UMNH group. Compared with the chloroform-methanol (CM) method, the diversity and concentration of volatile compounds were significantly reduced in NH and UMNH, and key volatiles such as hexanal, (E)-2-heptenal, and 1-penten-3-one were strongly associated with lipid oxidation. This study provides both theoretical foundations and practical approaches for high-efficiency, high-quality oil extraction from crustacean processing by-products.