Nature has a great open-access article on the technology of MRI neuroimaging, responsible for the majority of ‘brain scans’ that are used in medical examinations, scientific studies and media reports.
Understanding the technology of MRI scanners is not just of interest to medtech geeks, it is essential to be able to interpret and design brain imaging studies.
We tend to think that all brain scanners do the same thing, but comparing raw data across just two scanners can be a big problem.
Imagine if there were two different ways of measuring the layout of a room, one involving placing paper squares on the floor and another by tying string between all the objects. The conclusions might be similar but trying to merge all the raw results into one big data set would be a pain.
A similar problem affects researchers using brain scanners, because manufacturers may use different magnetic pulse sequences, different coils, different processing software and have to tweak the settings for each individual installation.
As a consequence, various studies are now developing ways (mainly data processing algorithms) to ensure that even relatively simple procedures, like structural scans, can be reliably compared across different scanners.
But even using just one scanner, it’s important to know what the technology is doing because this determines what aspects of the body’s physiology it measures.
The Nature article focuses on the latest developments in MRI scanner technology but also functions as a great brief guide to how scanner go from magnetic coils to measuring brain activity.