GE reveals plans in digital x-ray with debut of EG&G partnership

GE reveals plans in digital x-ray with debut of EG&G partnership

EG&G's large-area detectors are capable of real-time fluoro imaging

GE Medical Systems last month finally broke its silence regarding the development of direct digital detectors. The Milwaukee company announced that for the last three years it has been working on amorphous silicon-based digital detectors with EG&G Amorphous Silicon, and that it hopes to begin incorporating the detectors into its x-ray systems next year.

GE made the announcement at a press conference on Aug. 21 in Santa Clara, CA, where EG&G has built a high-tech manufacturing facility to assemble the detectors. GE executives in attendance, including GEMS president and CEO Jeffrey Immelt, said the event was the culmination of 10 years of work and over $100 million in R&D funding. EG&G has an exclusive contract to provide the detectors to GE, which will use them in a wide range of x-ray systems, including conventional x-ray, radiography/fluoroscopy, angiography, and mammography.

GE's announcement resolves long-standing questions that had been swirling about the direction that medical imaging's largest vendor would take in pursuing direct digital x-ray. The market potential for digital x-ray is enormous: Some 70% of the data in radiology is produced by analog-based x-ray systems, and existing technologies for digitizing these data have technical drawbacks. Until last month, however, GE had given few hints as to how it would approach the market.

Despite its low profile, GE actually began working on digital detector technology in the 1980s at its Corporate Research & Development center in Schenectady, NY. Several years ago, the company realized that it needed a partner to help develop the detectors, and began cooperating with EG&G Amorphous Silicon, a division of high-technology conglomerate EG&G, Inc. of Wellesley, MA. EG&G has been working on amorphous silicon detectors for the past 15 years.

From the beginning, one of the greatest challenges to EG&G engineers was to manufacture amorphous silicon panels large enough for medical imaging applications. Other digital technologies, like those based on charge-coupled devices (CCDs), have small detector areas that require the use of fiber-optic lenses or reducers to downsize the image, which can introduce artifacts. There are also problems with amorphous silicon approaches: Other vendors developing amorphous silicon detectors must tile together multiple panels to produce a full-size image, a process that GE and EG&G believe can introduce artifacts at the seams of the tiles. EG&G is capable of producing arrays with active areas as large as 41 x 41 cm, without the use of tiling.

Another advance reached by EG&G is the digitization of dynamic studies. Some flat-panel developers report that their systems are years from being able to digitize dynamic exams like R/F or angiography; at last month's news conference, GE and EG&G displayed dynamic cardiac cath images digitized with the panels.

"We have two big advantages over our competition: One is the ability of EG&G to produce these very large plates, which means we can take the whole image at once, just like an ordinary x-ray film," said Lonnie Edelheit, senior vice president at GE's CR&D unit. "The other advantage is that we can do fluoroscopy real-time, as well as high-quality radiography at the same time. That adds a lot of value."

The detectors are manufactured in a highly automated, 74,000-square-foot EG&G plant, using technology similar to that employed in making semiconductor chips, but on a much larger scale. Each EG&G amorphous silicon array is based on a glass substrate and consists of a cesium iodide scintillation layer, which captures x-rays and converts them to light; below that is an array of amorphous silicon photodiodes, which take the light and convert it into an electrical signal. The signal is read through a matrix of thin-film transistors, and is sent to a workstation for viewing.

EG&G is capable of producing 41 x 41-cm arrays with pixel sizes as small as 100 microns, but the need for such resolutions may be limited to full-field mammography. Conventional x-ray panels will have 200-micron pixel sizes, and both types of panels are capable of 16-bit dynamic range.

Mammography will be the first GE modality into which the detectors will be incorporated. GE has built the amorphous silicon detectors into a Senographe DMR mammography system, which is contained in a mobile mammography van that can be used for telemammography applications. Exams are acquired in remote locations and sent via T1 satellite feed to a central facility for interpretation.

The detectors will be incorporated into other x-ray systems after mammography. The panels will definitely be used in new x-ray systems, but GE has not decided whether the panels will be available as retrofits to existing systems in the field. Two exceptions are the Senographe DMR and Senographe 800T mammography systems, which will accept retrofits. GE plans to enter the detectors in the Food and Drug Administration 510(k) process at the beginning of 1998, and to begin feasibility studies of the detectors in the same time frame.

GE's development time line puts it on a collision course with Sterling Diagnostic Imaging, the company that until now has been the most recognizable proponent of direct digital detector technology. Other firms with positions in the direct digital market include Trixell; dpiX and Varian; Canon; and Optical Imaging Systems.

The impact of the detectors on the PACS market could be massive. Existing technologies for digitizing x-rays are cumbersome, and this is believed to be one of the impediments to more widespread adoption of PACS. GE's digital detectors will also provide a boost to GE's PACS efforts. The ability to provide direct digital acquisition of x-rays across a wide spectrum of radiographic applications should spur an already rapidly growing PACS market, according to Anthony Lombardo, global general manager for GE's Integrated Imaging Solutions division.

"This will be the final stimulus for the (filmless) department, because a large part of PACS is oriented towards small applications that are using (computed radiography) in the ICUs or ERs," he said. "The main department's conversion is going to be the real key."