From spectral signals to spatial maps: Precise prediction and visualization of PAHs in roasted lamb using an...

Application of combined indicators in quantitative analysis is relatively uncommon. This study introduced the entropy weighting method (EWM) combined with fluorescence hyperspectral imaging (FHSI) to develop a polycyclic aromatic hydrocarbon (PAHs) composite index (PCI). Using generalized and heterogeneous two-dimensional correlation spectroscopy, characteristic fluorescence peaks of trace components were resolved and the sequence of spectral changes was tracked. With PCI as the external perturbation condition, using PCI as the outer winding condition, the fluorescence intensity of PAHs varied sequentially from 465 nm to 442 nm, 433 nm, and finally 420 nm, suggesting that the energy level transition of pyrene occurs prior to that of fluoranthene during the roasting process. The CARS-PLSR model achieved an R²_C of 0.9472 and R²_P of 0.9350, with RMSEC and RMSEP values of 1.4998 μg/kg and 2.0710 μg/kg, respectively. Model performance was satisfactory, and other modeling approaches also yielded R² values above 0.7. Furthermore, spatial visualization of PCI in roasted lamb was achieved, confirming the accuracy and reliability of the proposed method for PAHs content prediction.