pca (1)

Dutch researchers used a new handheld hyperspectral imaging system to obtain information about nutmeg powder samples in the wavelength region of 400–1000 nm. The samples used to develop the method were 15 authentic samples, seven adulterant materials (i.e. 1 pericarp, 1 shell, and 5 spent samples) and 31 retail samples. Furthermore, another set of adulterated nutmeg samples were artificially prepared by mixing authentic material with spent powder (5–60%). Principal Component Analysis (PCA), Partial Least Squares-Discriminant Analysis (PLS-DA) and Artificial Neural Network (ANN) models were applied to the spectral data to construct the models, and authenticate the retail samples. The PCA showed successful spatial separation of authentic samples from adulterant materials. The ANN model predicted and showed the ability to detect adulteration at levels as low as 5% of added product-own material, which was more accurate than the PLS-DA model. Microscopic analysis was applied for comparison with hyperspectral imaging and to verify possible sample modification. It was concluded that method has good potential for the development of a visual quality control procedure for nutmeg powder authentication.

 Read the abstract here

Read more…