IBM and the Mayo Clinic on Wednesday announced the opening of a new collaborative research facility designed to speed the processing of compute-intensive algorithms used to deliver medical images from patients’ MRI and CT scans.
The Medical Imaging Informatics Innovation Center (MI3C) is the latest product of a four-year partnership between IBM and the Mayo Clinic, giving radiologists and physicians access to IBM’s latest blade hardware and Cell microprocessor architecture — first used in Sony’s PlayStation 3 — to render 3-D medical images that used to take hours in a matter of minutes.
“There have been algorithms out there for years that can take two-dimensional images and process them into 3-D but they take a lot of processing power and can take hours to run,” Bill Rapp, IBM distinguished engineer and chief technology officer for IBM’s Healthcare and Life Sciences team, said in an interview with InternetNews.com.
“But radiologists often don’t use them because they only have five to 10 minutes to do a reading. By putting all this computing power together, these images can be registered 50 times faster and improve the quality of radiology reading and productivity,” he added.
Rapp said IBM and the Mayo Clinic are sharing the costs of staffing the six-person MI3C staff while IBM provided “hundreds of thousands of dollars” worth of hardware. For this investment, IBM gets the opportunity to see what features and applications have the greatest commercial promise for future product releases and medical researchers get access to the technology and the ability to showcase to garner more federal research grants.
In practice, the high-speed computations allow radiologists to fill in the imaging gaps and blurs created when patients breath or make slight movements during examinations. If multiple scans of a particular organ or limb are slightly different—say the head is tilted a bit to the right or the left in each scan—the algorithm adjusts to provide a consistent image.
The MI3C is located on the Mayo Clinic’s campus in Rochester, Minn. and is open to third-party researchers and medical professionals.
