Taste profiles are crucial for the market success of black garlic, yet investigations into its physicochemical constituents closely related to taste quality remain limited. Additionally, the development of prediction models for taste sensory quality has not been extensively explored. This study aims to develop a cost-effective colorimetric sensor array (CSA) capable of simultaneously and quantitatively predicting the key physicochemical constituents and taste sensory quality of black garlic, integrated with chemometric algorithms. A multi-channel taste visualization sensor array was designed based on pH indicator color changes, indicator displacement assay (IDA), and silver nanoparticles. To establish quantitative prediction models, chemometric algorithms including partial least squares regression (PLSR) and support vector machine regression (SVR) were employed. The results revealed that nonlinear SVR models outperformed linear PLSR models in predicting reducing sugars, amino acid nitrogen, total acid, and the taste sensory attribute, achieving correlation coefficients for prediction (Rp) of 0.9863, 0.9232, 0.9666, and 0.9170, respectively. These findings highlighted the potential of integrating CSA with appropriate chemometric strategies as a reliable and promising approach for monitoring dynamic changes in key physicochemical constituents during black garlic processing and assessing the taste sensory quality of black garlic and similar food matrices.
Read full abstract