The main thing I was doing at work this week was helping for a couple of days on a training event for a study that involves ultrasound of patients with Giant Cell Arteritis (GCA), a condition that can affect various arteries, particularly but not only the the temporal arteries that run just in front of either ear. Untreated, it can lead to serious consequences such as permanent loss of site and ultrasound has been shown to be a quick and effective way to check for it. While further investigations, such as a temporal artery biopsy, may be required, a positive ultrasound scan is sufficient to confirm clinical diagnosis of the condition. The key sign is being able to spot a ‘halo’ around the arteries and my particular role this week was setting up and looking after a test using videos and still images but I also joined the group of volunteers allowing trainees to practise on them to get used to picking out normal arteries.

It got me thinking about how ultrasound works. A limitation of ultrasound compared to biopsies and to other imaging modalities is that it is relatively low resolution although recent decades have seen great leaps forward and halos, if present, can now be measured to a fraction of a millimetre. You can increase the resolution by increasing the frequency although modern equipment only goes up to about 20MHz. One of the limitations, as explained towards the end of this informative page, is that higher frequencies penetrate less deeply. That set me pondering as to why? After all, X-rays are much higher frequency still yet pass right through the body.

Fortunately, there were plenty of sonographers around to ask. The explanation that made sense was that, unlike X-rays, where you pass them through the body and record on a film on the other side, ultrasound is about capturing the signal that bounces back. It wasn’t a complete explanation – none of them were physicists (sonographers are more often physicians, interested in curing patients, than geeks who want to know how things work) – but good enough for now.

Oh, and I took the test myself. It was a relatively fair attempt as I’d been relying on the categorisation of the images from experts so I didn’t have them memorised. I got 90%, which is a pretty sound result, although I haste to add that is a very narrow demonstration of knowledge. Set loose on an ultrasound machine, I suspect I’d struggle to identify the right arteries although I should be fairly good at identify halos if I did!