At its simplest, mass spectrometry imaging turns a tissue section into a molecular map: no labels, no antibodies, and thousands of species recorded at once. The principle is straightforward. What makes it demanding is that the usual analytical safety net, a separation step before the mass spectrometer, is absent, so the instrument must resolve a very complex mixture on mass alone, at every one of many thousands of positions.
Key Takeaways
|
The Core Idea of MS Imaging
Rather than extracting analytes into solution, the instrument samples the tissue surface directly at a known position, ionises whatever is present there, and records a spectrum. The stage moves, the process repeats, and after some thousands of positions every pixel in a grid holds a complete mass spectrum. Selecting any mass-to-charge value and plotting its intensity across that grid produces an ion image.
The consequence worth internalising is that the acquisition is untargeted by construction. You are not detecting a chosen analyte; you are recording everything that ionised and choosing afterwards what to examine. A review of mass spectrometry imaging for spatially resolved multi-omics mapping in npj Imaging makes the point that label-free MSI can image thousands of molecules in a single experiment without prior knowledge and without labels or antibodies — a genuinely different proposition from a technique where you decide the target list in advance.
Working in analytical science?
Register for a FREE Separation Science account to subscribe to the Separation Science Newsletter.
Three properties of the resulting dataset follow directly, and they shape everything downstream. The data is hyperspectral rather than pictorial, so it is a cube of intensities across two spatial dimensions and one mass dimension. It is sparse and uneven, because ion yield varies with local tissue composition. And it is large, because the spectral dimension is retained at every position rather than collapsed into a single value.
How Do MALDI, DESI, and SIMS Compare?
These three do most of the work in practice, and they differ in the physics of how material is removed from the surface and ionised. That single difference propagates into resolution, analyte range, and how much the sample must be manipulated beforehand.
| MALDI | DESI | SIMS |
Desorption mechanism | Pulsed laser energy absorbed by an applied matrix, which transfers charge to analytes | Charged solvent droplets strike the surface and pick analytes up into secondary droplets | Primary ion beam sputters material directly from the surface |
Ionisation hardness | Soft, with minimal fragmentation. Produces mainly singly charged ions | Soft | Hard, fragmenting larger molecules |
What limits lateral resolution | Laser spot size and matrix crystal dimensions | Spray footprint and lateral spreading of solvent | Primary beam focus, the least constrained of the three |
Reported resolution | Commonly single-digit to tens of microns | Typically coarser than MALDI | Roughly 50 nm to 10 microns |
Section after analysis | Consumed at ablated positions | Largely preserved, near-native state | Consumed, and depth profiling is possible |
Preparation burden | Matrix application is the critical step and a genuine craft | Minimal, which avoids a class of preparation artefacts | Minimal coating, but must be vacuum compatible |
Analyte range | Widest: metabolites, lipids, peptides, proteins, glycans, drugs | Metabolites, lipids, some proteins | Elements, inorganics, small organics |
Table 1. The three approaches compared on mechanism and practical constraint. Figures are as reported in the peer-reviewed and society literature; confirm current performance for any specific instrument with its manufacturer.
A society-published beginner's guide to mass spectrometry imaging sets these distinctions out in more detail, including the variants developed to work around each limitation. Nano-DESI, for instance, uses two capillaries to form a continuous liquid bridge at the surface, which tightens the sampled area considerably compared with a conventional spray.
Two practical notes for method selection. MALDI’s dominance is partly a matter of spectral simplicity, since producing mainly singly charged ions yields less crowded spectra than electrospray-based approaches, which matters when there is no chromatography to spread the mixture out in time. And SIMS occupies a distinct niche rather than competing directly: its resolution advantage is real and substantial, but hard ionisation fragments larger molecules, so the gain in spatial detail is paid for in molecular information.
Spatial Resolution Against Molecular Coverage
Resolution is the specification most often quoted, and the one most often misread, because in imaging it is bought with something. Two separate costs are worth distinguishing, since they are frequently conflated.
The first is sensitivity, and it is simple arithmetic: a smaller pixel samples less material, so fewer molecules are available to ionise at that position and low-abundance species drop below detection. The second is coverage, which is subtler. As the sampled volume shrinks, the species that continue to be detected are increasingly the abundant and readily ionised ones, so the observable molecular range narrows even where signal remains. High-resolution images of a handful of abundant lipids are considerably easier to obtain than high-resolution images of a broad metabolite panel.
Why Pixel Count Punishes Ambition Spatial resolution enters the time budget quadratically rather than linearly. As the FT mass spectrometry imaging literature puts it directly, reducing the pixel raster size by twofold, for example from 50 to 25 microns, results in fourfold more mass spectra collected. That single relationship governs experimental design more than any instrument specification. An acquisition that takes three hours at 50 microns takes roughly twelve at 25 microns over the same area, before any allowance for slower per-pixel acquisition at higher mass resolving power. It is also why the analyser question in the next section is not academic: at high spatial resolution, seconds per pixel become days per section. |
The practical discipline is to specify the resolution the biological question requires rather than the resolution the instrument permits, then establish what coverage is achievable at that setting on your own tissue. That sequence is worked through in Spatial Resolution vs. Sensitivity in MS Imaging: The Fundamental Trade-off.
Which Mass Analyser Suits Imaging?
This is the instrument decision that most shapes what an imaging experiment can conclude, and it is decided by a constraint peculiar to imaging. In a liquid chromatography experiment, the separation resolves much of the sample’s complexity before anything reaches the analyser. In imaging, there is no separation step at all.
The consequence is that the analyser must resolve the mixture on mass alone. A 2024 Analytical Chemistry study of ultrahigh-resolution MSI states the case plainly: because of the lack of a separation step prior to ionisation and the immense diversity of biomolecules, including numerous isobaric lipid species, coupling ultrahigh mass resolution to imaging is one way that complexity can be resolved at the spectrum level. That study reached over one million mass resolution across the lipid range of roughly 600 to 950 daltons, and noted that until then such platforms had been restricted to FT-ICR instruments.
Set against that is the pixel-count arithmetic above. Imaging therefore demands high resolving power and high acquisition speed simultaneously, and the available analysers trade one for the other.
Analyser | Resolving Power | Acquisition Speed | Imaging-Specific Consideration |
TOF | Good, though ultimate resolving power is typically several times lower than the FT analysers. Multipass designs reach far higher | Highest. Well suited to large areas and fine rasters | In-spectrum dynamic range depends on acquisition time, so pushing spectral rate narrows it |
Orbitrap | Very high, and recently demonstrated above one million in the lipid range when externally coupled to a high-performance acquisition system | Moderate. High-field instruments are considerably faster than earlier designs | Higher field allows a given resolving power in a shorter transient, which matters most at fine rasters |
FT-ICR | Highest of any current analyser | Lowest. At extreme resolving power, on the order of one spectrum per second | No automatic gain control over ions entering the cell per pixel, and ion yield varies with tissue, so pixel-to-pixel mass shifting from space charge is a documented risk |
Quadrupole and ion trap | Lower | High | Suited to targeted imaging of known species rather than untargeted discovery |
Table 2. Mass analysers assessed for imaging rather than in general. Performance figures are as reported in the cited literature and vary substantially by instrument generation and configuration.
The FT-ICR row deserves expanding, because it describes a failure mode that does not arise in infusion or chromatographic work. Work on 21-tesla FT-ICR mass spectrometry imaging notes that the parameters varying with magnetic field strength, including dynamic range and mass accuracy, matter especially in imaging, both because ion yield changes with tissue type and because there is no control over the number of ions entering the analyser cell at each pixel. Related work has documented the resulting pixel-to-pixel shifting of mass peaks from space-charge effects, which degrades mass accuracy in the averaged spectrum and undermines confidence in assignments. Mitigations exist, including lock-mass calibration during acquisition and post-acquisition recalibration, but the point for method development is that the highest-resolving instrument is not automatically the most reliable one for imaging.
On the Orbitrap side, work on Fourier transform imaging using external data systems observes that acquisition time can be reduced at higher magnetic field because high resolving power becomes achievable with shorter time-domain acquisitions — an advantage that matters most precisely at high spatial resolution, where the pixel count is largest. The general characteristics of the FT analysers relative to TOF are set out in an Analytical Chemistry review of Orbitrap mass spectrometry.
Ion Mobility and Tandem MS in Imaging
If the absence of chromatography is the central problem, then adding a separation dimension back is an obvious response, and this is where much recent instrument development has gone.
Ion mobility separates ions by their collision cross-section as well as their mass, which distinguishes species that are isobaric but structurally different. That is directly relevant to lipid imaging, where isomeric and isobaric overlap is the dominant identification problem. The ultrahigh-resolution study cited above notes that a MALDI quadrupole time-of-flight instrument with trapped ion mobility enabled an increase in peak capacity of more than 250 percent during ion mobility experiments — a substantial gain in the ability to separate a crowded spectrum, achieved without slowing the raster.
Two further capabilities are worth understanding when specifying an instrument.
- Tandem MS on tissue. Fragmenting a selected precursor at a position gives structural confirmation rather than mass alone, which raises annotation confidence considerably. It costs acquisition time, so it is typically applied to selected positions or targets rather than across a whole image.
- Data-dependent and data-driven acquisition. Deciding during the run which positions or masses warrant deeper interrogation makes better use of a fixed time budget than treating every pixel identically. This is an active area and is where the practical gains in information per unit time are likely to come from.
The trade-off to keep in view is that every added dimension consumes either time or ions, and in imaging both are strictly limited by the pixel. An instrument specification that looks generous in infusion mode can be constrained in imaging for exactly that reason.
How Do You Choose an MSI Technique?
Work from the analyte and the required resolution, in that order, rather than from the instrument.
If You Need | Start With | Because |
Broad untargeted coverage across molecular classes | MALDI, on a high-resolution analyser | Widest analyte range, and mass resolution substitutes for the missing separation |
Lipid identification with isomer discrimination | MALDI with ion mobility, or ultrahigh resolving power | Isobaric overlap is the limiting problem, not sensitivity |
Minimal sample manipulation, or a section you must preserve | DESI | Ambient operation avoids matrix application and leaves the section largely intact |
Subcellular or elemental detail | SIMS | Only approach reaching well below the micron scale, accepting the analyte constraint |
Large areas or many samples at moderate resolution | TOF-based platforms | Acquisition speed dominates the time budget at scale |
Confident annotation of a small target set | Any platform with tandem MS capability | Fragmentation raises confidence beyond accurate mass alone |
Table 3. Matching requirement to approach. In practice, most laboratories have access to one platform and adapt the question to it, which makes knowing the constraints more useful than knowing the ideal.
That last observation is worth stating honestly. Few laboratories choose an imaging platform from a blank sheet; most have access to one instrument and need to know what it can and cannot support. Understanding where your available approach sits in Table 1 and Table 2 is therefore more immediately useful than knowing which configuration would be theoretically optimal.
This section takes each approach further, working through matrix chemistry, laser parameters, and the practical craft of MALDI in MALDI Imaging Mass Spectrometry: How It Works; ambient operation, its artefact advantages, and the nano-DESI and liquid extraction variants in DESI and Ambient Ionization Imaging; and primary ion beams, cluster sources, and depth profiling in SIMS and High-Resolution Elemental Imaging. For where mass spectrometry imaging sits alongside the antibody and sequencing-based spatial techniques, see Spatial Analysis in Analytical Science: Mass Spectrometry Imaging and Spatial Omics. Further coverage across separation and detection methods is collected in our Omics topic hub.
This article was produced under Separation Science's AI Editorial Guidelines.
