University of Washington enhances Equator's BSP chips with computing functions to propel medical imaging

Campbell 04 April 2002The video processing power of Equator's Broadband Signal Processor (BSP) family of chips, which can perform up to 40 GOPs (billion operations per second), is propelling advances in medical imaging devices. Equator Technologies is a provider of video streaming and processing engines and platforms. Using Equator's chip, the University of Washington's (UW) Imaging Computing Systems Lab developed a range of image computing functions that enable Equator's software-programmable chips to support the specific requirements of various medical imaging modalities.


Equator's platform is ideally suited for today's medical imaging devices, which require high accuracy, a high level of overall image quality, and real time processing of substantial amounts of data. Equator's superior image and video processing capabilities, combined with its high-level C programmability, allow developers to lower the cost and reduce the time to market for a variety of telemedicine applications and medical imaging systems, including ultrasound, computed tomography (CT), magnetic resonance imaging (MRI), digital mammography, fluoroscopy, and digital radiography.

Traditionally, medical imaging devices have relied on the processing power of hardware-based ASICs, creating static devices and significantly lengthening the time it takes to incorporate new technology developments and deliver new advanced devices to market.

The UW Imaging Computing Systems Lab has developed 129 highly optimised functions which include arithmetic functions, morphology, spatial/frequency-domain filtering, segmentation, contrast enhancement, 3D volume manipulation and rendering, as well as geometric manipulations and transformation. With the library running on the BSP family, medical device manufacturers will have all the necessary tools to create the most advanced devices on the market. The flexibility of Equator's platform enables developers to further enhance and customise applications, allowing them to keep pace with the rapidly evolving advances in the medical field.

"When we decided to explore medical imaging, it quickly became clear that we would need a powerful programmable processor which would be capable of handling the many complex algorithms and data flows needed in each medical imaging modality", stated Dr. Yongmin Kim, professor and chair of bio-engineering at the University of Washington. "Equator's BSP processor was the only engine fast enough to support our high-level image computing requirements, and its software programmability ensured that new codes and algorithms could be quickly and easily developed."

"The medical imaging space is one of the most demanding with regard to image and video processing, and our BSP chip family is a perfect match for handling the complex algorithms that enable medical devices to produce very detailed and precise images for better diagnosis. These capabilities can provide insight into the body and its systems that might have previously required more invasive procedures", noted Dr. Avi Katz, president and CEO of Equator Technologies. "Not only are we enabling device manufacturers to achieve a faster time to market with new products, but the advances that we are helping to further will allow medical professionals to provide patients with world-class care."

Equator Technologies Inc.'s family of BSP chips is a unique, fully functional, software-programmable system-on-a-chip solution designed for the development of platforms for demanding video applications, from "Pixels to Packets". Equator's compiler technology delivers 100 percent C programming, enabling the rapid introduction and field upgradeability of new applications and devices. In addition to the chips designed for the medical imaging space, Equator will provide its customers with access to the complete range of functions developed at the University of Washington. A complete list of functions with performance figures is available at the MAP University of Washington Image Computing Library Web page.

Leslie Versweyveld

[Medical IT News][Calendar][Virtual Medical Worlds Community][News on Advanced IT]