Optimizing high-moisture meat analogue textures through Artificial Intelligence: The effect of sorbitol in soy protein concentrate blends

Abstract

This study explored the effects of sorbitol on the textural properties of soy protein concentrate-based high-moisture meat analogues (SPC-HMMA) using a novel approach that combines artificial intelligence (AI) and genetic algorithms (GA) to replicate the textures of chicken and beef, aiming to develop customized meat analogues with tailored textural properties. This method allows for the simultaneous adjustment of multiple parameters, effectively capturing the complex non-linear interactions between ingredients and processing conditions during extrusion. SPC with varying sorbitol concentrations and moisture levels was extruded under optimized screw speeds and temperatures. Texture profile analysis (TPA) revealed that hardness values decreased from 3893 ± 308 g at 0% sorbitol to 421 ± 54 g at 20% sorbitol while cutting strength values ranged from 5951 ± 544 g crosswise at 0% sorbitol to 1754 ± 134 g at 20% sorbitol. Moisture content played a significant role in the textural properties of the SPC-HMMA with lower moisture yielding harder and chewier analogues. Scanning electron microscopy (SEM) revealed alterations in the microstructure while FTIR spectroscopy and deconvolution analysis indicated significant alterations in protein secondary structure. Cooking yield increased from 142.56 ± 1.5% to 168.54 ± 2.12%, water absorption capacity increased from 329.41 ± 5.16% to 464.67 ± 5.28%, while oil absorption capacity decreased from 120.84 ± 1.89% to 93.39 ± 1.82% with increasing sorbitol levels. • Meat analogues (MA) from soy protein were supplemented with sorbitol. • Sorbitol affected the hardness and cutting strength in MAs. • Water absorption and cooking yield optimized at moderate levels of sorbitol. • AI methods were employed to optimize textures akin to chicken and beef.

Topics & Concepts