Illustration of effect of computer-aided detection (CAD) of breast cancer in mammograms.
Image Courtesy Jose Galvez |
Breast cancer is the most common cancer in women, and mammograms are one of the main tools doctors use to diagnose this potentially fatal disease. The rate of misdiagnosis from mammography is high, however—estimates range up to 30%—due in part to physical differences across patient populations, differences in mammography equipment and procedures and difficulty in using computers to help detect changes in breast tissue. The MammoGrid project applies grid technology to aid in the accurate detection of breast cancer using mammograms.
With MammoGrid, doctors can harness the computing power of the grid to run advanced algorithms on digital versions of mammograms. This type of computer-aided detection of breast cancer, when used in tandem with the traditional method of visually screening mammograms, may help increase the accuracy of diagnoses. Computer-aided detection is even more effective when used with large mammogram databases, something the project is helping to create. MammoGrid also enables hospitals and doctors to share image data, making it easier for women to get second opinions and to involve their personal physician in decision-making.
"In some countries where mammography may take place in a hospital far from the small village where the patient and her doctor live, this kind of connectivity could greatly enhance the quality of medical care," says Jean-Marie Le Goff, group leader of CERN's Technology Transfer Group.
MammoGrid began in 2002 as a three-year project funded by the European
Commission's Framework Programme 5. When the first phase ended, a new
collaboration led by the CERN Technology Transfer Group was formed to
determine if the MammoGrid prototype could be turned into a commercially
viable product, and to disseminate the technology to the poorly developed
Extremadura region of Spain.
"The new MammoGrid partnership includes some of the original MammoGrid
partners, the Spanish company Maat GKnowledge, and several institutions and
hospitals from Extremadura," says Jose Galvez, the MammoGrid project leader.
The collaboration will install an upgraded MammoGrid system at five hospitals within the next six months. Three hospitals in Extremadura, one in Italy and one in the UK will receive a MammoGrid setup that includes one specially-designed workstation and a "grid box" with the MammoGrid software. The software is an optimized mix of the gLite middleware developed by the EGEE project and the G distributed database system developed by Maat GKnowledge, which manages the metadata associated with mammograms.
"The physician or technician will use the MammoGrid acquisition system to drag and drop digital mammograms from the scanner to the grid," explains David Manset from Maat GKnowledge. "The MammoGrid workstation then takes them into the virtual environment of the grid, where they operate on the images with virtual tools—applying whichever algorithm is appropriate for that image. The workstation interacts with the grid box to seamlessly access the heterogeneous and distributed grid resources."
Learn more at the MammoGrid Web site.
—Katie Yurkewicz
e-mail this article
|
