Step 2: Acquiring the images

The plan for image acquisition was pretty simple. After some quick scans to plan everything (we always do that in MRI), we were going to acquire a nice 3D data set for the "fly through" and surface renderings in the video, and finally two takes of James singing with the track played though the earphone. In total, James should have spent less than half-an-hour in the scanner. This is a fly through the 3D data set (for those interested it is a 3D T1 weighted sequence with 1mm isotropic resolution):

All went to plan until we finished the first take and James told us that he found it very hard to sing as he could hardly hear the track through the headphones. We tried to push the sound volume to the maximum, use the in-room ceilling speakers, no success! Finally, we took the decision to put the Hi-Fi speakers in the room. In theory not a good idea as (a) you can severely degrade the image quality; (b) possibly not the best thing to do to your speakers as they work with a magnet! Thankfully, our magnet room is big enough to put the speakers far away from the scanner not to dammage them, and luckily, we did not have any problem with the images. You can see our relief when we got it right in the next video. If you pay attention, you can see the real time images of James singing in the bottom right corner of the screen:

With everything sorted out, we did two takes of the full song. Below is a full data set acquired by the scanner, those images are not edited (if you really want to know, we used a steady state sequence at >10frames/second). It is not a very exciting video (no sound) but you can see James swallowing before he starts singing, and towards the end of the movie:

There was also time for this faux-pas before the final take: