wasn ’ t the only advantage . “ When we first developed them , we actually thought that the benefit would mostly reside in the high signalto-noise of these coils ,” Rosen said . “ And indeed , the signal-to-noise is unprecedented . More recently , we ’ ve come to understand that there is another benefit , which , in the long run , might be even more important . And that ’ s that we can use those high-density coils to greatly accelerate the acquisition of the MRI data .”
Such accelerated acquisition , of course , would enable much quicker scan times .
“ GOBrain grew very organically out of this work ,” said Larry Wald , Director of the core facility for MRI research in the Center and a Professor of Radiology at Harvard Medical School . “ We ’ ve been pushing parallel imaging and improving things by factors of two and three in the clinic . We just came to realize that this could all be put together into a very fast , very efficient brain scan .”
The scanning protocol used with GOBrain incorporates five key pulse sequences employed in the majority of routine brain exams with MRI . Limiting it to these five helped to reduce the scan time to near where the developers wanted it to be . Still , it wasn ’ t quite enough to get the total time to within the five minute target .
To address this , they deployed a decidedly low-tech solution : In the typical MRI exam , the technologist will check in with the patient between scans , to see how he or she is doing , explain what is next , etc . But because the brain exam was now so much shorter overall , the technologist no longer needed to talk to the patient between scans . Removing this step of the process helped to shave off the necessary additional time .
Not Just Faster , Better
GOBrain provided the dramatically reduced scan times that the MGH radiologists had asked for , enabling higher throughput and increased efficiency for the hospital and less anxiety and discomfort for the patient . But the developers discovered that it had another , entirely unexpected benefit : “ When we looked at the images , we often found , somewhat to our surprise , that [ they ] were actually significantly better ” than those acquired with conventional scanning protocols , Rosen said . This was somewhat counterintuitive ; “ you would think if you spend more time acquiring the data it ’ s going to take a better picture .” But as it happened , there was a simple explanation .
One of the most common challenges with MR imaging is how to deal with motion artifacts — that is ,
The Martinos Center ’ s Thomas Witzel
anomalies in the image that result from the subject or patient squirming or otherwise moving around during the scan .
It turns out patients can generally hold still for the five minutes it takes to acquire images with the GOBrain protocol . “ Most patients are not very comfortable with being in a machine for a long time , so they will usually cooperate for the first five to ten minutes and then they will become unruly and start moving more and more ,” said Thomas Witzel , the Director of the Human MR Imaging Core at the Martinos Center , and one of the key players in the development of GOBrain . “ So the faster you get all the images the better the quality , because the patient moves a lot less in the beginning of the session .”
The GOBrain application is now in use at Massachusetts General Hospital , where it is living up to its promise of broadly improving MRI brain exams . It is commercially available from Siemens Healthcare .