IBM researchers and a team of doctors are building a database of digital images that they hope will enable oncologists to diagnose and treat cancer patients faster and with more success.
Researchers at The Cancer Institute of New Jersey have digitized CAT scans, MRIs and other images using a high-performance Power6 570 Series system and computational time on the World Community Grid, also known as the world's largest public computing grid. IBM donated the hardware and computational time to the center.
Digitizing images should enable doctors to diagnose cancers earlier and detect their growth or shrinkage more accurately during treatment, according to Robin Willner, vice president of the global community initiatives at IBM.
"Right now, the doctor is basically eyeballing it when he's analyzing tissues and biopsies," Willner told Computerworld. "They're trying to figure out what type of cancer it is and if there's been progress during treatment. If you digitize the image, you're able to compare numbers because you've turned an image into bits and bytes. Now it's a much more accurate comparison."
Researchers have been using the grid, which is run by IBM and is using Windows-, Mac OS X- and Linux-based software, to convert hundreds of thousands of images of cancerous tissues and cells into digital images. Once the images are digitized, the grid can check the accuracy of the digital information -- ensuring that the bits and bytes are translating into real diagnoses.
The World Community Grid acts like a virtual supercomputer that is based on thousands of volunteers donating their unused computer time.
"If we can improve treatment and diagnosis for cancer, that's great for everybody," said Willner. "There couldn't be a better use for the grid."
Now that the digitization has been checked, Willner said the next phase of the project is to build a database that will hold hundreds of thousands, if not millions, of these images. Using a US$2.5 million grant from the National Institutes of Health (NIH), The Cancer Institute of New Jersey, Rutgers University and cancer centers around the country will pool their digital images in the database.
Willner said the database will enable doctors to compare patients' new images to ones already in the database to help them diagnose the cancer and figure out the best way to treat it.
Doctors should be able to use the database to personalize treatments for cancer patients based on how other patients with similar protein expression signatures and cancers have reacted to various treatments.
"The overarching goal of the new NIH grant is to expand the library to include signatures for a wider range of disorders and make it, along with the decision-support technology, available to the research and clinical communities as grid-enabled deployable software," said David J. Foran, a director of The Cancer Institute of New Jersey, in a statement. "We hope to deploy these technologies to other cancer research centers around the nation."
This isn't IBM's first foray into the medical arena by any means.
Earlier this month, IBM announced that it had teamed up with the Mayo Clinic to develop a research facility aimed at advancing medical imaging. Researchers from both the Mayo Clinic and IBM are working at the new Medical Imaging Informatics Innovation Center in Rochester, Minn.
Bradley Erickson, chairman of radiology at the Mayo Clinic, said a joint team is already working to find ways to use the Cell chip, mostly known for running inside the PlayStation 3 video-game console, in a medical imaging system. Erickson said that the technology could reduce work that now takes minutes down to a matter of seconds, or that now takes hours to minutes.