Astronomers can get a lot of information about stars and galaxies not just by looking at the light coming from them but by "listening" to them as well. In the 20th century astronomers began building radio telescopes in order to tune into the signals coming from distant objects like pulsars (extremely compact stars spinning very quickly and beaming out radio signals) and galaxies. Giant dishes were built all over the world; in the UK the radio telescope at Joddrell Bank is probably the most famous example.
The signals collected by radio astronomers can be assembled into an image: a radio-map showing where powerful and energetic processes are taking place in the object being studied.
The images collected by radio telescopes suffer from blurring and distortion just as images taken with ordinary optical telescopes do. During the 1960s and 1970s astronomers and mathematicians came up with impressive techniques to improve the quality of the images so that more detail could be extracted and perhaps saving the cost of getting better pictures by building a bigger and better telescope. When digital imaging evolved in the 1980s and 1990s ordinary backyard astronomers with smaller telescopes were looking to get more out of their pictures too. For example, one of the tricks employed is to take lots of images of an object and then use a computer to combine them to get a better picture.
Medical radiography is the process by which the internal structures of the human body can be imaged using a source of x-ray radiation. The problems faced by radiographers taking x-ray images are similar to those encountered by astronomers looking deep into the universe. However, where an astronomer might take lots of images of a galaxy and then combine them together to give a better final image, a radiographer can’t do the same. Repeatedly exposing a patient to x-rays to get a better picture is not an option; a good picture has to be taken first time and using the lowest radiation dose possible. Radiographers have to walk a fine line to do this. Too much radiation is bad for the patient (but the x-ray image will be good) and although less radiation is preferable, the image quality will be poorer. The radiographers are also constrained by the x-ray machines themselves. Producing x-rays puts an enormous heat load on the machine components and so the exposure time must not be too long. This could be avoided by illuminating the patient using a larger x-ray source - but that would blur the image to an unacceptable level.
In applications such as mammography, radiographers are typically trying to see tiny microcalcifications less than one tenth of a millimetre in size. Microcalcifications can be often be benign or they can sometimes be the first sign of a process leading to breast cancer; x-ray images need to be as clear and sharp as possible to see microcalcifications. The images produced by radiographers come after a careful balance of factors relating to patient safety and image quality, but even so, some blurring remains.
It's almost ten years since my PhD at Northumbria University ended. For the best part of four years I'd been involved in studying the problems of image quality that I’ve described. The Regional Medical Physics Group at Newcastle General Hospital allowed me to take x-ray images of test objects using a decommissioned machine. I was able to adapt and apply some of the image processing techniques used by radio astronomers to improve x-ray images and make microcalcifications easier to spot. I also showed that the radiation dose to the patient could be reduced without sacrificing image quality.
Having studied image processing techniques I still use sometimes use them on my own pictures of the Moon and planets taken through my telescope with a digital camera. It's s also gratifying to see that the same mathematics can be used to solve problems connected with seeing distant astronomical phenomena as well ash the very small components of our own bodies.
Welcome to my blog!
Dr Adrian Jannetta. Amateur astronomer, maths teacher and science enthusiast.