Manufacturing challenges snarl production of 1T open MR scanners

Siemens and Philips target late 2003 for commercialization

High-field open MR scanners are technological phantoms, existing in spirit but not in substance. While GE Medical Systems and Hitachi are in production with systems they describe as high-field opens, neither meets the traditional 1T benchmark of a high-field scanner. Siemens and Philips are each developing 1T open scanners, but neither has gone into production with a commercial version--nor is it likely that either company will do so any time soon.

At the annual meeting of the International Society for Magnetic Resonance in Medicine in May, executives from the two companies admitted they are struggling to overcome challenges in the design and production of these systems. While acknowledging the problems, however, each sought to minimize their significance.

"We have the technology understood," said Robert Hebel, Siemens director of MR product management. "The problem is on the cost side."

Siemens has one clinical system in operation and two others slated for installation over the next few months. The company does not expect to enter mass production with Rhapsody for at least another 18 months, however.

"We are currently reworking the product," Hebel said.

Philips has a 1T open magnet operating in an engineering bay at its R&D facility in Best, the Netherlands, said Peter Luyten, Ph.D., senior field director of Philips MRI business development. But before a clinical system can emerge, the company must resolve a number of technological issues. Many of these relate to the unique design of the system and, particularly, the magnet that drives it.

"The geometry of an open system is completely different (from that of a closed system)," Luyten said. "Going to a high-performance level is a big step. And that is what we are working on as expeditiously as possible to get (the Panorama 1.0T) in the field."

The first clinical installations will not occur until 2003, he said, and full-scale production is about 18 months away.

The apparitions of high-field open systems first appeared when GE unveiled its Signa OpenSpeed in mid-November 1999. Although OpenSpeed was billed as a high-field open, the system was designed to run at 0.7T, well below the traditionally accepted high-field benchmark of 1T. This spurred Siemens Medical Solutions to rush a mock-up of its 1T open to the 1999 RSNA exhibit floor, promising a "true" high-field open product. A year later, Siemens showed a much-refined version of the scanner, called Rhapsody, at the 2000 RSNA meeting. Philips showed one of its own, the Panorama 1.0T.

During the 2000 RSNA show, Siemens executives predicted that Rhapsody would be installed at its first clinical site in summer 2001, with routine deliveries in the U.S. before the end of 2001. Philips was more conservative, forecasting commercialization by the end of 2002. Since then, it has become apparent that the path to a 1T open system has met more difficulties than expected.

Siemens' problems are related primarily to its open magnet, which is being made by England-based Oxford Magnet Technology, Hebel said. (Siemens owns a majority equity stake in the magnet manufacturer.) The rest of the technology on Rhapsody is well established. Much of the hardware and software in use on Siemens' high-field Magnetom products have been ported to Rhapsody.

"Technically, we have no problem. The systems out there are running and they perform quite well--we meet our gradient specifications and homogeneity requirements. We can do fat sat, we can even do spectroscopy," Hebel said. "The difference is only with the cost of the system. How we have developed it makes the cost of production higher than expected."

These problems raise questions as to whether a commercially viable 1T system can actually be produced. To be competitive, a 1T open product must not cost much more than a conventional closed high-field system. Despite increasingly sophisticated technology, the price of conventional high-field systems has steadily dropped to as low as $1.5 million or even less, depending on options.

GE appears to be under little pressure to come up with a 1T open product. The company is solidly behind its OpenSpeed, which Dennis Cooke, GE's general manager of global MR, routinely describes as gaining more and more customer interest. Cooke expects interest to grow with each passing month that the long-promised 1T systems remain absent from the marketplace.

Philips, through its purchase of Marconi Medical, has the Panorama 0.6T (formerly known as the Marconi Infinion HFO), which the company rolled out in March at the European Congress of Radiology in Vienna. Philips kicked off its North American marketing effort for the 0.6T system at the annual meeting of the American Society of Neuroradiology in May. (The scanner is priced around $1.4 million in the U.S. to compete with GE's 0.7T OpenSpeed and Hitachi's 0.7T Altaire.) Philips is also positioning its ultracompact Infinion 1.5T at the high end of its "open" product line.

Siemens, however, has no such fallback. Company executives are pushing forward, meeting commitments to site clinical systems, while trying to resolve financial challenges inherent in its current design. One Rhapsody is operating at a medical facility in Dortmund, Germany, producing what Hebel called "awesome images in all applications from head to toe." The system handles between 10 and 15 patients daily.

"We didn't have any installation problems, and we don't expect any to occur in the future," he said.

A second Rhapsody will soon be installed in Florida, and a third is scheduled for a California site by the end of summer. The design of these systems, however, while technically proficient, is not cost-effective. Hebel explained that the Rhapsody magnet design is particularly complex.

"It is not a normal tunnel system," he said. "The magnet is really two systems and it is, therefore, more complicated, more difficult to build."

Philips engineers are focusing on problems that cross a "whole spectrum," according to Luyten. Included are the Panorama 1.0T's gradients and radio-frequency coils.

"It is not one item," he said.

The challenges being addressed by these two companies underscore the fact that 30 months after the debut of a work-in-progress 1T open system, a commercial version is still at least a year and a half away. If this schedule is met, four years will have passed from the first showing of a 1T to its commercialization.