The introduction of computed tomography has contributed to major advances in medical imaging research and diagnosis. CT imaging has taken plain x-ray radiographs from 2D flat images to 3D images. In 2D imaging 3D objects appear flattened – the image from objects in the front overlap with those in the back. CT on the other hand takes a series of images around a sample, which are then processed by a computer to produce slices through the sample. This 3D view shows the structure more clearly and also shows objects inside others. This 3D view shows the imaged structures more clearly, it also illustrates the relationship between adjoining structures, distinguishing between objects that are beside each other or within each other. Colour CT has the potential to be as revolutionary to medical imaging as the move from 2D images to 3D. The addition of a colour spectrum to the spatial resolution provided by traditional 3D CT contributes significantly more information; with colour CT it is now possible to determine what an object is made of.
Visible light is made up of many different wavelengths, called a spectrum. We see these wavelengths as different colours, or all the wavelengths together as white. Different things appear as different colours depending on which wavelengths they reflect (colour), and how many photons of each wavelength they reflect (brightness).
Photography provides a good analogy when considering the impact of moving from black and white images to colour. Black and white cameras detected how much white light (all wavelengths of visible light) is being absorbed or reflected by objects. However they do not distinguish what is happening to the different wavelengths within this light. Colour photography measures three different wavelengths – red, yellow and blue. These wavelengths are reflected or absorbed off objects in view, and the camera is able to record this. The mix of reflection of red, blue and yellow gives objects their colour. The image to the left demonstrates how much more information is contained in a colour image than black and white.
Similarly, x-rays are also produced in a range of wavelengths called a spectrum. These different wavelengths are not visible to, or distinguished by, the human eye but they are able to be detected with film and digital devices. Standard radiographs use x-rays over a wide spectrum. They measure how much of the x-rays are attenuated by the object they pass through. Attenuation is a reduction in the total intensity of the x-rays as they pass through an object. They cannot measure how each different wavelength is being individually attenuated (this is equivalent to being able to measure the brightness but not the colour). The Medipix3 detector is able to measure how specific energies of x-rays are being attenuated. While these are outside of the range of wavelengths the human eye can see, they are ‘true colour.’
All materials attenuate the various wavelengths differently. This is because the atomic structure of materials is different. For example, on a regular x-ray image bone (predominantly calcium) highly attenuates the x-rays and appears white. Iodine, a common contrast agent, also highly attenuates the x-rays and appears white. However, across the spectrum the two materials are behaving very differently. Therefore during processing of the colour CT image, these two materials can be separated. The image below is of the same sample. In the colour image the back bone is green (calcium) and the stomach (filled with iodine contrast) is yellow. In the black and white image these two objects both appear white.
The Medipix3RX chip is the detector used in the MARS scanner. It is able to distinguish both density and atomic variation in a sample. The density determines the brightness of the image and the atomic structure (the fundamental materials of which the sample is made) determines the colour.
This leads to two possibilities: if the attenuation of a material is known then what would have previously been an unknown in an image can be determined. The attenuation of many things is known, such as calcium and iodine, so if these appear in an image they can be easily identified. If the image is comprised of all unknowns, then these unknowns can be separated out, despite not knowing what they are.