Fusion of food profiling data from very different analytical techniques

The increasing demand for safe, authentic and high-quality food requires high-resolution and rapid analytical methods to reliably verify these properties. As a result, a variety of approaches based on different analytical techniques, e.g. based on mass spectrometry, spectrocopy or imaging, have been developed. However, these approaches often focus on specific aspects of the complex composition of food and thus only consider a small part of food properties. In order to gain a comprehensive understanding and to obtain powerful approaches for food testing, it is particularly advantageous to combine data from very different analytical techniques. The combination of datasets with different properties in particular poses challenges and there are different approaches for their fusion. In this article, we analyse and evaluate the current state of the art for fusing very different food data from various analytical techniques and make recommendations for approaches that can usefully be applied to data fusion.