Computational Modelling for Optimization of Thermosonicated Sohshang (Elaeagnus latifolia) Fruit Juice Using Artificial Neural Networks

Abstract

The study involved subjecting sohshang (Elaeagnus latifolia) juice (SJ) to thermosonications (TS), a process integrating ultrasound and heat, with a range of independent variables. Specifically, three explored distinct amplitudes (30%, 40%, and 50%) alongside three temperature settings (30°C, 40°C, and 50°C) and four treatment durations (15, 30, 45, and 60 minutes) were used in the experiment. A variety of quality parameters were analyzed such as antioxidant activity (AOA), ascorbic acid (AA), total flavonoid content (TFC), total phenolic content (TPC), yeast and mold count (YMC), and total viable count (TVC). Thermosonicated sohshang juices (TSSJ) successfully achieved highest content of AA (69.15 ± 0.99 mg/100 ml), AOA (72.93 ± 1.62%), TPC (122.03 ± 4.23 mg GAE/ml), and TFC (116.14 ± 3.29 mg QE)/ml) under ideal circumstances. Also, minimal TVC and YMC in these juices have been observed. The best results for AA and TFC were observed at 40°C with 40% and 50% amplitude over processing times of 45 and 60 min. To optimize the extraction processes with various objectives, artificial neural network (ANN) was established with an original experimental planning methodology. Overall, the investigation demonstrated that TS is an effective method to enhance the nutritional and microbiological qualities of sohshang fruit juice. The use of ANN in the optimization process is particularly valuable in achieving desirable outcomes. As the food and pharmaceutical industries seek natural and bioactive substances, TSSJ holds great potential for various applications.