Thursday, November 3, 2011

New Algorithm Speeds up MRI Scans

New Algorithm Speeds up MRI Scans

Magnetic Resonance Imaging (MRI) scanners are used daily by doctors to produce images of a person’s body, organs, and structure. These images are examined to find problems such as internal bleeding, tumors, infections, cancer and many other harmful things that can’t be diagnosed by from looking at a person’s exterior. MRI scanners use radio waves and magnetic fields to create multiple images of the body. Although these scans are very useful, patients need to lie inside the MRI machine for about 45 minutes without moving. Researchers, scientists, and engineers from MIT’s Research Laboratory of Electronics have developed an algorithm that will cut this time down to only 15 minutes.

An Algorithm is an “artificial intelligence system that mimics the evolutionary, survival-of-the-fittest process to generate increasingly better solutions to a problem (Hagg 111).” This new algorithm will allow for quicker scans by using previous scans as a starting point for new scans. This allows the scan to be based of something, rather than starting with nothing. The difficulty in creating this algorithm is determining what information from the scan will be useful for following scans. By having an outline to work from, the algorithm will make the scans run for less time. A simple example to this complex algorithm would be writing a paper. If you make an outline that you follow every time you write a paper, your paper will be completed quicker and more efficiently. If you dive into the paper without a foundation, it will take longer and will probably be unorganized.

Algorithms are optimizing systems in the sense that they use various inputs to generate optimal outputs for a given task. Unlike writing a paper, an algorithm for such important and high tech MRI scanners is extremely complex. Elfar Adalsteinsson, an associate professor of electrical engineering, explained, “If the machine is taking a scan of your brain, your head won't move from one image to the next, so if scan number two already knows where your head is, then it won't take as long to produce the image as when the data had to be acquired from scratch for the first scan (Medical News Today).” The algorithm will take inputs from the first scan to create boundaries for the next. It determines all the next information it will need to use to create the next image, and then examines what prior information can be used.

MIT’s researchers will need to strive towards tweaking this algorithm to make it as efficient as the 45-minute scans are. The MIT team said that by making MRI’s quicker, they are losing some quality. This is because they don’t want the scan to gather too much information from the first scan because it could lose vital and unique information that different scans can pick up. Although people want their MRI’s to be finished 30 minutes earlier, most would prefer a scan that won’t miss crucial information. I think this will be MIT Research team’s biggest obstacle, but by continually working to perfect the algorithm, a 15-minute MRI scan can be available in the near future.


Haag, Stephen, and Maeve Cummings. Management Information Systems for the Information Age. New York: McGraw-Hill Irwin, 2010. Print.

No comments:

Post a Comment