Conductus and Stanford to develop low-field MR mammography scanner

Partners hope to build prototype next year

A group of scientists from Stanford University have hooked upwith MRI technology developer Conductus to build a dedicated MRmammography scanner that could cost a fraction of the price ofexisting MRI systems. To achieve their goal, the partners willemploy a novel prepolarized MRI (PMRI) technique and high-temperaturesuperconducting receiver coil technology.

Imaging vendors that are developing dedicated MR mammographyscanners include Advanced Mammography Systems and InnerVisionMRI. Other companies employ breast coils for use in standard MRImachines. Most of these firms, however, state that their devicesare intended for follow-up to x-ray mammography rather than forroutine breast screening.

The problem is that MR mammography for screening applicationsis too expensive to compete with conventional mammography. Low-fieldMRI scanners could bring the cost of MR mammography down to apoint where it could be competitive, but image-quality issuesrelated to the low signal-to-noise ratio (SNR) of low-field systemsremain an obstacle.

Conductus of Sunnyvale, CA, and the Stanford group are attackingthis problem by integrating low-field magnets with superconductingtechnology in an effort to improve SNR to a point where low-fieldMRI breast screening is both clinically useful and economicallyfeasible.

The PMRI system under development at Stanford's Magnetic ResonanceSystems Research Laboratory will use two inexpensive pulsed electromagnetsinstead of the more expensive superconducting magnets found inconventional MRI scanners. To maintain high image quality withultra low field strength, the PMRI system will use low-noise cryoelectronics,including superconducting receiver coils and cryogenic amplifiersof the type now being developed by Conductus.

"We believe that the marriage of our two technologies canproduce an MRI mammography scanner comparable in cost to today'sx-ray mammography," said Albert Macovski, a professor ofelectrical engineering and radiology, who is credited with developmentof PMRI.

Macovski and research associate Steven Connolly co-direct theStanford PMRI program. By operating with two relatively smallelectromagnets, one for prepolarizing hydrogen protons and onefor the readout of data, the PMRI system might be configured asa C-arm device with an aperture that is roughly the size of theanatomy to be imaged, Macovski said.

"The PMRI magnets can be highly inhomogeneous and stillwork quite well, because we separate the functions of polarizationand readout that current systems cannot do," he said.

"We pulse the low-field PMRI magnet up to a high field strength(such as 0.5 tesla) for a fraction of a second and then turn itoff. Even though we read out at a very low field, it's the polarizationoperation that determines image quality and scan time."

As a result of PMRI technology, the SNR of the system dependson the strength of the prepolarizing pulse, not the strength ofthe main magnetic field, said Frank Muenneann, a Conductus projectengineer. The field strength of the main magnet at rest is 0.02tesla.

Cooling image quality. The high-temperature superconducting technologydeveloped by Conductus is designed to improve system performanceand raise SNR by lowering the noise floor, as opposed to raisingthe signal level. That can be done using superconducting technologyto eliminate resistance -- and therefore noise -- in radio-frequencycoils, according to Randy Simon, vice president of technologyprograms at Conductus.

"This makes it possible to reduce scanner field strengthwithout sacrificing SNR," Simon said. "

In low-field MRI situations we're looking at (a ratio of) approximately5:1 in terms of signal-to-noise enhancement with our superconductingcoil technology. That's really the payoff, particularly if you'regoing to start losing image quality because you've lowered scannerfield strength in the first place."

Another California firm, Superconductor Technologies of SantaBarbara, has also developed superconducting RF coils designedto improve SNR in low-field MRI scanners (SCAN 5/4/94).

Conductus opted for breast imaging as an initial applicationfor its technology for two reasons. It allows the use of smallmagnets, which are cheaper. In addition, breast tissue has relativelylow conductivity, thus generating little noise.

Furthermore, recent clinical studies indicate that MRI mightbe able to detect breast cancer at an earlier stage of developmentthan can x-ray mammography, without the use of ionizing radiation.The goal of the Conductus/Stanford effort is to commercializea PMRI system that would cost about $200,000 to $300,000, Simonsaid.

"In principle, that should be possible, because the onlysizable magnet in the system is the prepolarizing magnet, whichis very crude compared to the superconducting magnets in conventionalscanners," Simon said.

Conductus and Stanford hope to have a working prototype of thesystem built in a year.

There could be a major stumbling block looming on the horizon,however. The Conductus/Stanford project is funded by a $4.4 millioncontract granted to Conductus by the Naval Research Laboratoryand the Advanced Research Projects Agency (ARPA) under the Departmentof Defense's Focused Research Initiative.

Budget cuts to the Focused Research Initiative, which are beingconsidered by Congress, could cut off much of the federal moneybeing used to fund the project, thus stalling development of thesystem for years. The initiative's funding was initially set at$20 million. A version of the 1996 budget passed by the Houseof Representatives has proposed cutting that amount in half, whilea Senate version has proposed eliminating the funding entirely.