Grid computing to aid breast cancer research

From...

By Todd R. Weiss

(IDG) -- Breast cancer treatment and research are getting a big boost from a new grid computing system that will link hospitals, doctors and researchers as they seek new ways of fighting the disease.

The computing grid, which will initially link four hospitals in North America, was announced Wednesday by IBM as the latest in a growing list of grid computing projects the company has undertaken. The grid will use IBM servers and technologies to collect and store data for the project.

Dr. Robert Hollebeek, director of the National Scalable Cluster Lab at the University of Pennsylvania in Philadelphia, where the grid is being built, said the system is expected to bring about major advances in diagnosis and treatment for breast cancer patients.

The grid will allow doctors to store and instantly call up and compare digital images of patient mammograms, letting them spot changes in the images.

IDG.net INFOCENTER Computerworld main page Computerworld's topic-oriented communities Computerworld's IT career center Free daily newsletters Related IDG.net Stories Medical imaging strains storage IBM to build grid for gene research Grid computing gets push from Sun, IBM and Compaq Features PC World.com Product Finder Mastering e-commerce by degrees Fast-track training puts nontechies in IT jobs Visit an IDG site Choose a site: IDG.net CIO Computerworld Darwin Dummies.com GamePro.com The Industry Standard ITWorld.com InfoWorld.com JavaWorld LinuxWorld Macworld Online Network World Fusion PC World Publish.com UnixInsider IDG.net search

"If a patient comes in for an annual exam and something looks a little suspicious, if a previous year's record [or earlier] can be looked at, that's a really big advance clinically," Hollebeek said. "If you have it as a reference, and you can digitally compare mammograms as you go forward, that's a new capability you didn't have before."

For doctors and researchers, the growing grid will allow them to combine patient records from vast numbers of hospitals, providing important pools of new data in the search for breast cancer cures and treatments.

"It's like having little notes and check stubs laying around the house," Hollebeek said. "Sooner or later, you have to get organized."

The grid computing project, which began about two years ago, now includes the University of Pennsylvania, the University of Chicago, the University of North Carolina-Chapel Hill and the Sunnybrook and Women's College Hospital in Toronto. The project is being built in collaboration with a group from Oak Ridge National Laboratory and is being funded by the National Library of Medicine.

The first four hospitals will begin using the grid in January, Hollebeek said. "We're very focused on making this clinically useful as soon as possible," he said. "It has huge potential to be useful to a lot of people."

David Turek, IBM's vice president of high-performance computing, said the project will help bring together vast quantities of information to improve breast cancer treatment. "This is just a terribly important area of health care," Turek said.

The main parts of the grid computing system will be housed in data storage facilities set up around the nation as the grid grows, Hollebeek said. The goal is to link some 2,000 hospitals and oncology clinics in the next few years, using two servers set up in each location as a "portal" linked to the nationwide grid.

Plans call for the eventual construction of a 1-petabyte (1 million-gigabyte) database of digital patient information for breast cancer research and treatment.

The grid will store mammograms in digital form and provide analytical tools to help physicians diagnose individual cases and identify cancer "clusters" in the population. It will also give authorized medical personnel near-instantaneous access to patient records and reduce the need for expensive film X-rays, which also take up lots of physical storage space.

The grid will use a three-tier architecture built from IBM eServers and open protocols from the Globus Project, a nonprofit open-source grid computing effort.

Each participating hospital will have a portal built from two IBM eServer xSeries systems. One will be used as a temporary repository for the digital data, while the other will link the grid to the next generation of the Internet, called Internet2. The data will then be sent to a metropolitan hub -- an IBM eServer Cluster 1600 Unix system.

As the grid grows, data from several metropolitan hubs will be funneled to a high-capacity regional hub, which is now being prototyped with an IBM eServer 1300 Linux system. The system will use IBM's AIX Unix, Linux and Microsoft Windows.