Cocoa powder surface composition during aging: A focus on fat

Markedly Enhanced Analysis of Mass Spectrometry Images Using Weakly Supervised Machine Learning

Supervised and unsupervised machine learning algorithms are routinely applied to time-of-flight secondary ion mass spectrometry (ToF-SIMS) imaging data and, more broadly, to mass spectrometry imaging (MSI). These algorithms have accelerated large-scale, single-pixel analysis, classification, and regression. However, there is relatively little research on methods suited for so-called weakly supervised problems, where ground-truth class labels exist at the image level, but not at the individual pixel level. Unsupervised learning methods are usually applied to these problems. However, these methods cannot make use of available labels. Here a novel method specifically designed for weakly supervised MSI data is presented. A dual-stream multiple instance learning (MIL) approach is adapted from computational pathology that reveals the spatial-spectral characteristics distinguishing different classes of MSI images. The method uses an information entropy-regularized attention mechanism to identify characteristic class pixels that are then used to extract characteristic mass spectra. This work provides a proof-of-concept exemplification using printed ink samples imaged by ToF-SIMS. A second application-oriented study is also presented, focusing on the analysis of a mixed powder sample type. Results demonstrate the potential of the MIL method for broader application in MSI, with implications for understanding subtle spatial-spectral characteristics in various applications and contexts.

Thermal properties and volatile compounds profile of commercial dark-chocolates from different genotypes of cocoa beans (Theobroma cacao L.) from Latin America

There is a growing interest in the identification of chemometric markers that allow the distinction and authentication of dark-chocolates according to their cocoa geographical origin and/or genotype. However, samples derived from Latin American cocoa, including specimens from North and South America, have not been studied in this context. An exploration of the melting behavior, fat composition, bioactive content, and volatile profile of commercial darkchocolates was conducted to identify possible patterns related to the genotype and/or origin of cocoa from Latin America. The melting properties were evaluated by DSC and related to fat content and fatty acids profile. Total polyphenol, anthocyanin, methylxanthine, and catechin content were analyzed. Finally, the volatile compounds were extracted and identified by HS-SPME/GC-MS and were analyzed through Principal Component Analysis (PCA) and the Hierarchical Cluster Analysis Heatmap (HCA Heatmap). The fatty acids profile showed a relationship with the melting properties of dark chocolate. The samples exhibited two glass-transition temperatures (T_g) at ≈19 °C and ≈25.5 °C, possibly related to traces of unstable polymorphic forms of monounsaturated triacylglycerides. The analysis of bioactive compounds demonstrated great variability among samples independent of the cocoa origin, genotype, and content. The PCA and HCA Heatmaps allowed discriminating against the chocolates in relation to the cocoa origin and genotype. Compounds like tetramethylpyrazine, trimethylpyrazine, benzaldehyde, and furfural could be considered as dark-chocolate aroma markers derived from Latin American cocoas (North American region). The 2-phenylethyl alcohol, 2-methylpropanoic acid, 2,3-butanediol, 2-nonanone, and limonene for derived from South America. And the 2-phenylethyl acetate, 3-methyl-butanal, and cinnamaldehyde could allow to distinguishing between regional genotypes.

Direct determination by portable ED-XRF of mineral profile in cocoa powder samples

The present study has exploited the rapidity of the analysis and the multi-elemental capability of the energy dispersive X- ray fluorescence (ED-XRF) technique for the mineral profile determination in cocoa powder. A fast, cheap and environmental sustainable method without reagent consumption or toxic waste generation has been proposed. The samples can be prepared in the form of pellets of 13 mm in diameter and 2-3 mm thickness. The different internal calibrations used by ED-XRF equipment did not provide accurate results when comparing the mineral profile with the concentration obtained by Inductively Couple Plasma Optical Emission Spectroscopy (ICP-OES) after microwave assisted digestion of samples. For direct ED-XRF analysis of the cocoa samples, an external calibration using as standards the cocoa samples diluted with sugar was prepared. The analytical parameters of Relative Standard Deviation and Limit of Detection for the determined elements are adequate to the concentration levels found in the samples.