Mass Spectrometry Imaging in Food: The Power and the Limits

LCGC International spoke to Laura Righetti of Wageningen University & Research and Wageningen Food Safety Research about mass spectrometry imaging (MSI) in food safety and authenticity, the subject of her RAFA 2026 talk, "Imaging Food: The Power and the Limits of Seeing Molecules in Food." MSI shows where a contaminant or marker compound sits within a food sample, without staining or labels, at a time when laboratories want to know why a hazard arises as well as how much is present. Ion mobility and multireflecting time-of-flight instruments now give MSI much of the molecular specificity that chromatographic separation provides in liquid chromatography–mass spectrometry (LC–MS). Even so, Righetti expects most food companies to outsource MSI to specialist laboratories for the next 5 to 10 years.

Your talk at RAFA 2026 is framed as “the power and the limits” of seeing molecules in food. In practical terms, what can MSI show us about a food sample today that no other single technique can?

Using MSI is like taking a picture of food with your smartphone, but instead of capturing colors, it captures molecules. It uniquely combines the molecular specificity of mass spectrometry (MS), telling us what molecules are present, with spatial information, showing where they are located within the sample. Unlike microscopy, this can be done without the need for staining or molecular labels, providing a direct molecular map of the food.

Your recent paper “Mass Spectrometry Imaging in Food Safety and Authenticity: Overcoming Challenges and Exploring Opportunities” notes that despite a surge in MSI publications, food safety and authenticity work remains largely “proof of concept,” with most studies still qualitative rather than quantitative.¹ What's the biggest technical bottleneck between a compelling proof-of-concept image and a validated method a food safety laboratory could actually deploy?

In my opinion, the biggest bottleneck is still quantification. MSI is inherently a pixel-based technique: We measure signal intensities at each pixel and use them to reconstruct the spatial distribution of a compound. However, converting the signal intensity of a pixel into an absolute concentration, and then relating that concentration to a regulatory limit for a contaminant, is far from trivial.²

For this reason, I think the translation of MSI into routine food safety testing is particularly challenging. In contrast, food quality and authenticity applications may be more readily achievable, especially when authentic and fraudulent samples are analyzed under identical conditions. In those cases, MSI can be used comparatively—for example, by evaluating the spatial distribution and relative abundance of specific molecular markers, without necessarily requiring absolute quantification.

This paper also describes a “location-to-function” paradigm—understanding not just where a toxin ends up but why, and what that means biologically. Using your work on aflatoxin B1 and its plant metabolite aflatoxin M₂ in maize as an example, how does knowing the localization change what a breeder or regulator should actually do?

Our aflatoxin work illustrates why localization matters.³ Knowing that a toxin is present is important, but knowing where it accumulates can help us understand the biological processes that drive contamination and detoxification. For example, the spatial distribution of aflatoxin B1 and its plant metabolite aflatoxin M₂ can provide insights into how the plant responds to the presence of xenobiotics and where detoxification mechanisms may be active.

For breeders, this information can help identify tissue-specific resistance or detoxification traits that could be strengthened through selection. For regulators, however, the main value is more indirect: Localization improves our understanding of contamination mechanisms and helps identify the most relevant biomarkers and target tissues for monitoring. In other words, moving from “where is the toxin?” to “why is it there?” brings us closer to understanding the biology behind food safety risks.

Ion mobility MS and next-generation MS instruments with ultrahigh mass accuracy are central to overcoming MSI’s current limits. Concretely, what becomes possible with this combination that wasn’t before—are there isobaric or isomeric food contaminant pairs, comparable to the malvidin/aflatoxin M₂ case your group has published, that this could finally resolve?

One of the fundamental challenges of MSI is that, unlike conventional liquid chromatography (LC)–MS or gas chromatography (GC)–MS workflows, there is no chromatographic separation before mass analysis. Everything is measured directly from the sample surface, which makes molecular identification particularly challenging in complex food matrices.

Our malvidin/aflatoxin M₂ case is a good example. These compounds share the same molecular formula and therefore the same exact mass, so mass accuracy alone, regardless of how good it is, cannot distinguish them. In that case, the key information came from MS/MS fragmentation patterns.

This is where ion mobility MS adds tremendous value. By introducing an additional gas-phase separation dimension before MS, compounds with identical masses can be separated based on their size, shape, and charge, even when they overlap in the mass spectrum. This significantly improves the ability to distinguish isobaric species and reduces chemical interference directly in MSI experiments. There are examples in the literature that have shown the separation of closely isobaric species such as PC(32:0) and PC(34:3), which differ by only 3 mDa.⁴

At the same time, multireflection time-of-flight (TOF) instruments can provide ultrahigh resolving power and sub-ppm, even sub 200 ppb, mass accuracy, allowing the separation of species that differ by only a few millidaltons. For example, the multireflection TOF has been shown to separate lipid species such as PC(33:1) from neighboring lipid species and interfering background signals. On lower-resolution instruments, these contributions would merge into a composite image, potentially leading to incorrect biological interpretation.⁵

So, rather than a single contaminant pair that suddenly becomes accessible, I see the major advantage as a substantial reduction in ambiguity. The combination of ion mobility and next-generation MR-TOF instruments brings MSI closer to the molecular specificity that chromatographic separations have traditionally provided, while preserving the spatial information that makes MSI unique.

You describe a future “food atlas,” modeled on projects like the Human Cell Atlas, mapping toxic and bioactive compounds down to single-cell resolution in crops. What would a food company or regulator be able to do with that resource that they can't do today?

Today, we mostly measure how much of a compound is present in a food sample. A food atlas would tell us where compounds are produced, accumulated, transformed, or detoxified, potentially down to the tissue or even single-cell level.

For food companies, this could accelerate the development of crop varieties with improved quality, safety, and nutritional traits by identifying the specific cells or tissues associated with desirable or undesirable compounds. For regulators, it could provide a much deeper understanding of contamination mechanisms, helping move from detecting hazards to understanding how and why they arise.

Importantly, these spatial maps could also help us trace the fate of nutrients, bioactive compounds, and contaminants after consumption, linking their distribution in foods to their distribution and metabolism in tissues,⁶ as well as being of value within the food itself. In that sense, a food atlas could help bridge the gap between food composition and health outcomes, connecting nutrient distributions in foods with how diet reshapes metabolism in the body.

More broadly, a food atlas would shift the focus from inventorying molecules to understanding their function in space and time. Just as the Human Cell Atlas transformed our view of human biology, a food atlas could help us understand the molecular architecture of crops and link location to function, quality, authenticity, safety, and ultimately nutrition and health.

Your paper suggests food companies will eventually need to either invest heavily in MSI instrumentation and specialized staff, or outsource to dedicated laboratories. Which path looks more realistic to you over the next 5 to 10 years, and what has to change for MSI to become routine rather than a research tool?

Over the next 5 to 10 years, I think outsourcing to specialized MSI laboratories is the more realistic scenario. While the instruments are still expensive and require specialized expertise, the bigger challenge is data analysis. MSI generates extremely complex datasets, and there are currently no widely accepted, standardized workflows for data processing, molecular annotation, and interpretation.⁷

For MSI to become a routine tool rather than primarily a research technology, we need not only more robust and user-friendly instrumentation, but also automated, standardized, and validated data analysis workflows. Ultimately, companies are interested in actionable answers, not just beautiful molecular images. The current situation is similar to where many omics technologies were years ago: extremely powerful but still requiring substantial expert intervention.

Until we reach that point, I expect most food companies to access MSI through specialized service laboratories, while larger companies may invest in in-house systems for high-value applications such as authenticity, quality, and product development.

For a laboratory or researcher who’s new to MSI and considering adopting it, what should they look out for going in—the mistakes people commonly make, or the assumptions about sensitivity, quantitation, or sample prep that trip up newcomers?

My first piece of advice would be: Don't think of MSI as just another mass spectrometry technique. What makes MSI powerful is the spatial information, but that also brings additional challenges.

One common misconception is about sensitivity. Just because a compound can be detected by LC–MS does not automatically mean it will be detectable by MSI at the same level. In MSI, you’re often working with much smaller amounts of material per pixel and without the benefit of chromatographic separation.

Another misconception is quantification. Beginners often assume that a brighter signal means a higher concentration. That’s not always true because signal intensities can be strongly affected by matrix effects, ion suppression, and local sample composition. MSI is excellent for relative comparisons, but absolute quantification remains challenging.

And finally, never underestimate sample preparation. In MSI, sample preparation is often the experiment. The quality of the tissue section, analyte preservation, matrix application, and experimental design can have a huge impact on the final image.

References

1. Tortorella, S.; Bartels, B.; Suman, M.; et al. Mass Spectrometry Imaging in Food Safety and Authenticity: Overcoming Challenges and Exploring Opportunities. Trends Food Sci. Technol. 2025, 155, 104803. DOI: 10.1016/j.tifs.2024.104803
2. Unsihuay, D.; Mesa Sanchez, D.; Laskin, J. Quantitative Mass Spectrometry Imaging of Biological Systems. Annu. Rev. Phys. Chem. 2021, 72, 307–329. DOI: 10.1146/annurev-physchem-061020-053416
3. Righetti, L.; Bhandari, D. R.; Rolli, E.; et al. Unveiling the Spatial Distribution of Aflatoxin B1 and Plant Defense Metabolites in Maize Using AP-SMALDI Mass Spectrometry Imaging. Plant J. 2021, 106 (1), 185–199. DOI: 10.1111/tpj.15158
4. Spraggins, J. M.; Djambazova, K. V.; Rivera, E. S.; et al. High-Performance Molecular Imaging with MALDI Trapped Ion-Mobility Time-of-Flight (timsTOF) Mass Spectrometry. Anal. Chem. 2019, 91 (22), 14552–14560. DOI: 10.1021/acs.analchem.9b03612
5. Cooper-Shepherd, D. A.; Wildgoose, J.; Kozlov, B.; et al. Novel Hybrid Quadrupole-Multireflecting Time-of-Flight Mass Spectrometry System. J. Am. Soc. Mass Spectrom. 2023, 34 (2), 264–272. DOI: 10.1021/jasms.2c00281
6. Vander Kooi, C. W.; Gentry, M. S.; Sun, R. C. Shaping Human Health and Nutrition Through Innovations in Spatial Metabolism. Annu. Rev. Nutr. 2026, 46, 49–73. DOI: 10.1146/annurev-nutr-062024-111053
7. Jetybayeva, A.; Borodinov, N.; Ievlev, A. V.; et al. A Review on Recent Machine Learning Applications for Imaging Mass Spectrometry Studies. J. Appl. Phys. 2023, 133 (2), 020702. DOI: 10.1063/5.0100948