The development of MEG source reconstruction methods through models of human retinotopy

点击翻译按钮获取中文摘要

Magnetoencephalography (MEG) is a brain scanning technique which allows one to make estimates of electrical activity in the human brain as it changes over milli-seconds. The technique has many advantages- it involves simply putting your head in a helmet, is extremely sensitive and has no adverse effects. The one major disadvange however is that we do not have all the information we need. That is, in order to say precisely where activity is taking place you have to make a guess about what is likely to be happening. It is the same problem we solve automatically when we go to the cinema. We only have two dimensions of information (the flat screen), and from that we make estimates of a three dimensional scene (a room with people in it). We are able to do this because we know things about the world, like the way shadows fall, and that people are all more or less the same size. The problem with imaging the brain of course is that we don't know much about it. We have been lucky however and found a rule that seems to give us reasonable brain images. This rule is that no two areas of the brain should be doing the same thing at the same time. In this grant we want to slightly change this rule and see what kind of an effect it has on our brain images. For example, we 'd like to allow bits of the brain that are right next to one another to do more or less the same thing. If the images get better (and we can judge this because we have other scanners that are slower but don't rely on a rule set) then we are on the right track. We will know more about how the brain works, and we will be able to produce millisecond by millisecond images of the brain working.