Abstract
Prolonged storage degrades the quality of Wenyujin Rhizoma Concisum (PJH), a functional food ingredient rich in volatile bioactive terpenes, calling for an investigation and a rapid non-destructive identification method. This study adopts a holistic "bioactivity-composition-sensory" approach to evaluate PJH over time, combining cell assays, high-performance liquid chromatography, electronic nose, and hyperspectral imaging. Results show that extended storage leads to marked declines in anti-inflammatory and antioxidant activities, along with reductions in key volatile terpenes, weakened aroma, and color fading. A machine learning model was subsequently constructed based on hyperspectral data for storage year discrimination with 100% accuracy. These findings systematically reveal the multidimensional quality deterioration of PJH and establish a scientific basis for determining its shelf life. This holistic perspective and hyperspectral machine learning approach in this study offer a paradigm applicable to quality monitoring and stability research of other volatile-rich functional ingredients.