A seed testing lab is asked two things at once: tell us whether this lot is good, and do not consume the lot finding out. Cutting seed open answers the first and fails the second: it shows the embryo, the fill and the damage inside, and ends that seed. Where a breeding line is a few hundred seeds, that cost is real.
X-ray imaging separates the two demands. The seed is imaged through its coat, graded on what the image shows, and can still be planted. That property moved cabinet X-ray into seed houses, gene banks and plant research groups. The real questions are what a radiograph decides, what it misses, and which system fits your seed.
What a seed radiograph actually shows
Seeds suit X-ray because the defects that matter are density differences inside an intact object. Empty and partially filled seed reads clearly against fully developed seed. Cracks from handling or hard drying appear even when the coat looks flawless. Voids, malformed embryos and insect infestation, including larvae inside the seed, show without cutting anything open.
Fill quality is the measurement most labs come back to, because it is comparable. Once imaging is configured for a species, fill reads across lots, treatments and breeding lines, and the comparison holds week to week. That beats one striking picture of one bad seed.
Where X-ray screening fits the testing routine
Germination testing answers the commercial question slowly. Days to weeks pass before the trays report back, and by then the lot may be in storage or in the ground. Imaging runs where the decision is taken: incoming lots, batch release, sorting a breeding population before planting.
The two fit together: a radiograph tells you which seed deserves a germination test, and which lot has a physical problem no germination test will repair. A batch that is half empty seed does not need a week of trays. Imaging is triage, not a verdict.
Imaging area and throughput decide which XSEED fits
The KUBTEC XSEED line comes in three sizes, chosen by batch size and detail. XSEED mini is the entry point: a 7 × 12 cm imaging area at 50 µm with a 20 to 50 kV source. The mid-range XSEED uses a 12 × 15 cm detector, four magnification levels and automated imaging an operator can override. XSEED Plus carries the largest field of view at 24 × 31 cm, a 50 kV, 2.0 mA source, the most sensitive detector in the line and five magnification levels up to 3×, for minute cracks and early infestation.
Field of view is not only about speed. It sets how a tray is laid out, how many seeds stay in a known position, and how much handling sits between sample and result. Merkel Technologies supplies the XSEED line in Israel, and the first question we ask is about batch size, not detectors.
Repeatable grading beats a single sharp image
All three models run DIGISEED analysis software, and its value is consistency: testing is configured per species and repeated identically, not left to whoever is at the system that morning. When the result feeds a breeding decision, a certificate or a rejected delivery, the difference between a method and an opinion is whether the next operator grades the same.
Image Blender helps too: it merges the X-ray with an optical view of the same specimen, so a finding sits on the actual seed and is easier to defend, teach and recheck.
Three questions seed X-ray imaging cannot answer
Viability is the first limit. X-ray shows structure, not life. A seed can be fully filled and free of cracks and still fail to germinate: dormancy, ageing, heat damage and physiological condition leave no reliable structural signature. A claim built on imaging alone is a claim about the seed, not the crop.
Chemistry and genetics sit outside the method. Seed treatment coverage, oil or protein content and genetic identity are separate analyses, and no contrast adjustment turns a radiograph into one of them. Pathogens are a related trap: the damage a fungus causes may show, but identifying the organism is different laboratory work.
Geometry is a limit too: a cabinet system produces a projection, so internal structure is compressed into a plane and overlapping seed can hide the feature being hunted. Where the interior of one seed must be reconstructed in three dimensions, following a crack through the endosperm or measuring a cavity, that is work for benchtop micro-CT, at far lower throughput. Labs whose imaging runs past seed into specimen radiography look instead at the same manufacturer's cabinet X-ray platform for the research laboratory.
Who operates the system, and what it needs
These are cabinet systems, meant to sit in a working laboratory rather than a dedicated radiation facility, and the mini can be moved on a cart. Planning is still needed. Radiation safety registration, operator training and a clear procedure for who images what belong before installation.
A gene bank with seasonal peaks and temporary staff has different needs from a group with two trained users. Before a purchase, the most useful step is a demonstration on your own species and defect list, the same instinct behind renting an instrument before committing to it.
The right seed X-ray system changes a decision
Seed imaging earns its place when it changes something. If images are collected, filed and never used to reject, sort or prioritise, the system is an expensive camera. If the grading turns back a suspect lot, or sorts a breeding population before planting, the value shows within a season.
The choice is less about the finest detector than about where the bottleneck sits in your own testing. Merkel Technologies works through that with seed producers, breeding programmes and research groups, starting with the seed the lab actually handles. The right system answers the question you keep asking, at the speed you keep asking it.