Analysis of Milk Microstructure Using Raman Hyperspectral Imaging.
Anastasiia SurkovaAndrey BogomolovPublished in: Molecules (Basel, Switzerland) (2023)
Optical spectroscopic analysis of the chemical composition of milk in its natural state is complicated by a complex colloidal structure, represented by differently sized fat and protein particles. The classical techniques of molecular spectroscopy in the visible, near-, and mid-infrared ranges carry only bulk chemical information about a sample, which usually undergoes a destructive preparation stage. The combination of Raman spectroscopy with confocal microscopy provides a unique opportunity to obtain a vibrational spectrum at any single point of the sample volume. In this study, scanning confocal Raman microscopy was applied for the first time to investigate the chemical microstructure of milk using samples of various compositions. The obtained hyperspectral images of selected planes in milk samples are represented by three-dimensional data arrays. Chemometric data analysis, in particular the method of multivariate curve resolution, has been used to extract the chemical information from complex partially overlaid spectral responses. The results obtained show the spatial distribution of the main chemical components, i.e., fat, protein, and lactose, in the milk samples under study using intuitive graphical maps. The proposed experimental and data analysis method can be used in an advanced chemical analysis of natural milk and products on its basis.
Keyphrases
- data analysis
- raman spectroscopy
- high resolution
- optical coherence tomography
- single molecule
- adipose tissue
- oxidative stress
- fatty acid
- healthcare
- high throughput
- white matter
- machine learning
- molecular docking
- magnetic resonance imaging
- density functional theory
- protein protein
- mass spectrometry
- convolutional neural network
- label free