Answers are grounded in this paper and cite their sources.
Fourier transform infrared spectroscopy is widely used for food quality assessment, but its performance depends on selecting spectral variables that capture chemical changes during processing. This study applied an unsupervised strategy to identify sensitive FTIR markers from the temporal evolution of spectral absorbance. Thermal degradation of commercial soybean oil heated at 180 °C for 72 h was used as the main case study, and the selected markers were compared with predefined wavenumbers, integrated spectral regions, and the full spectrum. Five highly responsive wavenumbers were identified and showed consistent trends during heating, capturing oxidation and structural changes more effectively than conventional approaches. Principal component analysis and K-means clustering indicated that these markers concentrated degradation-related variance into fewer dimensions and improved stage separation. The approach was further supported using three independent external FTIR food datasets. These results highlight evolution-sensitive spectral markers as compact, interpretable, descriptors of processing-related changes in food systems.
No takes yet. Share an insight, caveat, or question.
Marcos-Escobar et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: