The X-ray Color Revolution That Might Spot Hidden Tumors

Picture this: You have just twisted your ankle on a morning run. After hobbling to the doctor’s office and feeling the pain with each step, you are led into a dimly lit room. The technician positions your swollen ankle on a cold platform, disappears behind a protective barrier, and with a soft click, invisible rays pass through your aching joint.

The doctor calls you to his computer screen and shows you an image of your bones, seemingly unharmed.

“Good news, no fracture!”

You are relieved. But just as your pain medication kicks in, you find yourself staring at the greyscale image of your intact foot. A thought strikes you: How come X-rays are always black and white images and never feature even a hint of color? Clearly, it doesn’t have to do with the monitor; it’s happily displaying the doctor’s family vacation photos as a screensaver.

Also, when was the last time you saw a camera or smartphone that couldn’t take photos in color? In fact, smartphone manufacturers bombard us with more (often unnecessary) camera improvements every year. But for some reason, the images taken with a 100,000 $ hospital X-ray machine still basically look like the one Wilhelm Conrad Röntgen, the discoverer of X-rays, captured of his wife’s hand in 1895:

I understand that determining whether a bone is broken works perfectly fine using a black and white image — either there’s a gap or there isn’t.

However, X-rays also have far more sophisticated applications. When doctors need to identify cancer, bleeding in the brain, or other life-threatening conditions, they often turn to computed tomography (CT) scans. These involve capturing X-ray images from many different angles to reconstruct 3D information of the body’s interior. But in the end, we still get a greyscale image:

Clearly, that’s enough to tell organs apart and notice if there’s a suspicious lump where it doesn’t belong. But as it turns out, there are certain kinds of tumors that are really hard to detect with these scans: if the tumor has the same shade of grey as the normal tissue — which is a common occurrence with kidney, liver, and pancreas tumors — you are out of luck. Despite clearly not being the same thing, the tumor tissue and normal tissue are displayed as the same thing: blobs of grey.

Why isn’t there a way to get more information on the tissues out of these scans? The problem would be resolved and a lot more diseases might be easier to diagnose and treat early. Fortunately, a lot of research is currently underway with the goal of getting and displaying that additional information — by adding color to your X-ray. But to understand how this works, let us first have a look at the magic behind X-rays.

When Light Gets X-treme

Before we think about X-rays, let us first consider the light you encounter every day: what the sun or your desk lamp gives off is actually a bunch of electromagnetic waves. Think of them as a changing electric field, which creates a magnetic field, which then creates an electric field again — and so on forever (or until the wave hits your eye and ruins your perfect afternoon nap).

The most important property of these waves is their wavelength, which describes after what distance the wave repeats its pattern. All the light you are able to perceive with your eyes has wavelengths between 350 and 700 nanometers, which is comparable to a hundredth of the width of a human hair. If we take these (already tiny) wavelengths and make them about 10,000 times smaller, we end up with X-rays. This means X-rays are essentially the same phenomenon as visible light, just with wavelengths on the scale of an atom:

But why do X-rays mostly pass through your body while visible light doesn’t? The answer is related to the fact that waves with small wavelengths carry more energy: a lot more oscillations happen per distance traveled. This means that X-rays carry substantially more energy than visible light.

When light interacts with matter, it’s mostly interacting with the electrons surrounding the atomic nucleus. These electrons exist at specific energy levels, similar to steps on a staircase. For light to be absorbed, it (generally) needs to have precisely the right amount of energy to move an electron from one step to another:

Visible light excels at this process: In most of the atoms that make up your body, many energy levels match the energy of visible light. As a result, visible light doesn’t travel far before being absorbed by some atom, which unfortunately means humans are not see-through (though that would probably be quite fun).

The story is very different for X-rays: with their extremely high energy, they rarely match the typical energy levels found in our bodies. Consequently, they mostly pass through without being absorbed.

For an intuitive comparison, imagine jumping on a trampoline: if you jump at the natural bouncing rhythm the trampoline was designed for, it starts swinging wildly because you are transferring energy from your body to the jumping mat efficiently. This is similar to how the relatively slow oscillations of visible light interact with electrons.

However, if you jump extremely rapidly (like X-rays), your timing is so mismatched with the trampoline’s natural bounce that energy transfer fails and the jumping mat basically stays in place. Similarly, the electrons simply cannot “catch” the rhythm of these ultra-fast X-ray oscillations, allowing X-rays to pass through with minimal interaction.*

From Shadows to Images

But there are materials that even X-rays struggle to penetrate. Heavier elements, like those found in bones or metals, are more likely to absorb X-rays because they have more matching energy levels. This explains why calcium-rich bones are easy to see on X-ray images, while soft tissues like skin or muscle — primarily composed of lighter elements — are not.**

When you view an X-ray image, you are essentially looking at a shadow: you shine X-rays through the body, and in most areas, they pass straight through. Where bones are present, the rays are absorbed, less light reaches the detector and a dark spot is created in the image. Interestingly, for historical reasons, we invert these colors, making the dark areas appear bright (bone) and the bright areas appear dark (soft tissue or air)***.

Now that we understand the mechanism behind how the images are formed, it is clear why the image is in greyscale: we are measuring the amount of light intensity that gets through, which is either all of it (white), none of it (black) or something in between (grey).

There’s simply no way to get color information here. Case closed.

Or is there?

Unweaving the X-ray Rainbow

What does it actually mean to see colors? I told you earlier that visible light has a certain range of wavelengths. Your brain is capable of sensing each of these wavelengths and giving you the right color impression in your head:

If we apply the same principle to X-rays, that would mean we need a way to sense the wavelength of each X-ray wave that we measure. Sure, our brains still wouldn’t have an interpretation for those wavelengths, because we never evolved one (seems like the ability to detect X-rays with your eyes didn’t create too much of a survival advantage in the caveman era). But we could simply map each X-ray wavelength to a visible light wavelength that our brains can understand.

We actually already use this color-mapping technique for taking “photos” of galaxies and other interesting structures in space. We have telescopes capable of detecting many different wavelengths of light that humans can’t naturally interpret, including numerous wavelengths I haven’t mentioned in this article.

In an effort to create those impressive images for your desktop wallpaper, scientists associate each invisible wavelength with a corresponding color from the visible light spectrum. The result is a beautiful false color image:

But we can’t do this with traditional X-ray images. Here, we only measure the amount of X-rays passing through our object. It’s similar to placing a bucket outside to collect rain for an hour, then measuring the total water volume. This tells you how much it rained, but nothing about the size of individual raindrops. What we really need is a sensitive scale that measures each raindrop’s weight as it falls. This way, we still get the total volume (by summing everything up in the end), but we also receive information about every single drop.

This is precisely what photon-counting X-ray detectors do: they record individual light packets (photons). When an X-ray hits this type of detector, it creates a small electrical pulse whose size reveals its energy. The detector’s electronics then sort these pulses by size, essentially counting how many X-rays of each energy level passed through each point in your body.

Different tissues interact with X-rays of different wavelengths in unique ways, absorbing some wavelengths more than others. By detecting how tissues affect each specific wavelength, we can in principle capture a distinctive “fingerprint” for each material. We can later use this information to determine how much of each material is present in each pixel and display it using a color map.

Limited by the Technology of Our Time

Unfortunately, these detectors aren’t yet sophisticated enough at distinguishing wavelengths for this to work in clinical practice today, but early prototypes are being tested****. There is, however, a simpler technique — so simple that it’s already incorporated into most airport scanners:

The approach is called dual-energy X-ray imaging. Here, we scan an object with two different X-ray wavelengths in sequence, then combine information from both images to learn more about the materials present.

Think of it like photographing an object under different colored lights. If you take one photo using only green light and another using only red light, certain materials will respond differently to each. Plants, for example, would appear bright in the green-light image but remain relatively dark in the red-light photo, since they are very good at absorbing red light. You can then create a new image that indicates “how plant-like” each object is, based on how much it brightens when illuminated with green light.

You can repeat this for any number of materials: how metal-like, paper-like, or wood-like are the objects? Then you give each material a color code based on what it most closely resembles, as I showed you in the airport scanner example.

The same principle is also being applied to some CT scanners already: There is a condition called gout in which crystals form primarily around joints in the feet or hands, causing pain and inflammation for patients. Since we know the properties of gout crystals and how they should behave under different X-ray wavelengths, we can use a dual-energy CT to highlight all gout-like structures in green. This allows us to monitor how effectively the gout treatment is working:

Just to be clear: This is different from simply coloring the original grayscale CT image to make it more visually appealing, like in a paint-by-numbers book. We can actually determine how much of a specific material (e.g., gout crystals) is present in the scan with this dual-energy approach and assign that material an appropriate color.

The Colorful Future

Some things might never change. Twenty years from now, if you twist your ankle during another morning run, you will likely find yourself back at the doctor’s office. The X-ray they take will probably still be perfectly black and white — after all, why change a running system?

Yet, medicine evolves. It is of course too early to make any promises about what impact color-sensitive CT scans are going to have for sure. But from what we do know about first clinical results with dual-energy technology, they may help us identify abnormalities previously invisible to conventional scans. The ultimate promise of X-ray images in color isn’t just prettier pictures — it’s earlier detection, more precise diagnoses, and potentially life-saving interventions for conditions that would otherwise remain hidden until they’ve progressed too far.

What I love about the physics of X-rays is that it transfers so wonderfully into real-world applications that shape the medicine of the future — and hopefully end up making our lives better. The journey from Röntgen’s image of his wife’s hand to today’s emerging color-sensitive scans spans more than a century, all focused on a single goal: seeing what was previously invisible. For me personally, that’s enough to keep me showing up on the lab doormat every day.

*This is somewhat of an incomplete picture of X-ray interactions with matter, since X-rays aren’t only absorbed but also scattered into different directions. When an X-ray flies off in a very different direction, it essentially looks like it has been absorbed, since it won’t find its way to our detector anymore. What type of interaction is the most common depends on the energy of the X-ray and doesn’t lend itself to an intuitive and general explanation (at least one I could come up with).

**It is interesting to note here that water (of all things) is a counterexample to this. Water is quite transparent to visible light (for smaller thicknesses, think water glass rather than ocean). But it does show reasonable attenuation of X-rays, meaning visible light has an easier time passing through than X-rays. In fact, what we call visible light is the only part of the electromagnetic spectrum for which water is this transparent. Whether or not that’s a coincidence is a topic for another article.

***Before modern X-ray detectors, the images were captured using X-ray film, which remained white in unexposed areas (bone) while turning black in exposed areas (soft tissue or air). With digital detectors, we can display the image in whatever way we want — but the images were kept the same during the transition to not confuse radiologists.

****If you are interested in the details as to why: one of the problems is if a photon hits the detector too close to the border of two detector pixels. In this case, the energy deposited by the photon may be split up into two pixels and the detector mistakenly counts two low-energy photons instead of a single high-energy one.