A novel time-intensity data analysis approach: Hierarchical clustering heatmap coupled with proton-transfer-reaction mass spectrometry data in a non-spiked real food system

Time-intensity (TI) is a dynamic sensory evaluation method widely used to investigate the temporal evolution of sensory perceptions during food consumption. To better understand aroma perception, TI has been combined with instrumental analysis techniques such as proton-transfer-reaction mass spectrometry (PTR-MS) to interpret the relationship between sensory perception and in vivo aroma release in spiked model food systems. While spiked model foods are more readily investigated due to their known target volatile organic compounds and simplified matrix, studying non-spiked real food systems provides more practical insights into sensory perception and better reflects actual consumption conditions. However, investigating the relationship between TI and PTR-MS in non-spiked real foods remains challenging. Identifying aroma-related PTR-MS signals has been difficult due to the complexity of the detected data. Moreover, parameters extracted from TI and PTR-MS curves capture only limited aspects of the temporal dynamics, which can lead to incomplete interpretation of sensory-instrumental relationships. In this study, a novel analytical strategy combining the log 2 fold change (log 2 FC) method with hierarchical clustering heatmap visualization was developed to explore the relationship between sensory perception and in vivo aroma release in non-spiked natural yogurt. The log 2 FC method was used to screen aroma-related PTR-MS signals, while the heatmap visualization was applied to enable straightforward interpretation of the full temporal profiles of both TI and PTR-MS data. Our results reveal a correlation between screened PTR-MS signals and TI responses across panelists. Overall, this novel approach improves the interpretation of sensory-instrumental correlations in non-spiked real foods by providing an intuitive heatmap visualization. • Developed a novel log 2 FC-heatmap method for PTR-MS and time-intensity data. • Combined sensory and PTR-MS data to explore aroma perception dynamics. • Enabled intuitive visualization of temporal patterns via clustering heatmap. • Revealed strong correlation between time-intensity responses and PTR-MS signals. • Provided a novel statistical tool to study aroma perception in real foods.