NMR patent: a matter of infringement.

In 1974 Raymond V. Damadian, a physician, patented the design and use of a nuclear magnetic resonance (NMR) imaging device to scan the body in search of cancer. Since then, N,R scans -- now known as magnetic resonance imaging (MRI) scans -- have become a standard diagnostic technique employed by large hospitals and medical clinics throughout the world. However, only about 30 of the estimated 250 MRI machines in medical use have been made by Damadian's company, the Meville, N.Y.-based Fonar Corp. Since no one sought a license from Fonar to manufacture the device, Damadian reasoned that makers of those other machines much be infringing on his patent. He sued one such company. And in late November 1985, a jury in a Massachusetts federal district court case ruled that Johnson & Johnson Inc.'s Cleveland-based Technicare subsidiary--maker of 100 MRI medical devices--had indeed infringed upon Damadian's patent. But on Jan. 10, the judge overturned the finding of infringement, saying that, based on the evidence, the jury had rendered an invalid judgment.

The judge's ruling, which Fonar will appeal, seems to hinge on a scientific issue: Do physicians using NMR diagnose diseases--like cancer--based upon a comparison of the time it takes excited hydrogen nuclei in diseased and normal tissue to "relax" back to their ground state?

Although the spin orientation of atomic nuclei in the body is usually random, those spins will line up in a north-south polar orientation when the nuclei are acted on by a magnetic force. In NMR, such a magnetic force is applied to body tissue. Then a radio wave signal with a frequency that will be absorbed by hydrogen nuclei is beamed at the tissue. Nuclei absorbing this energy become excited, altering their alignment. When the radio signal stops, the nuclei relax back to their previous magnetic-field alignment. MRI devices measure the time it takes excited nuclei to relax, and display with a smilar contrast regions having the same relaxation time. These relaxation times are known as T1 and T2 (based on orientation--transverse or longitudinal -- of the magnetic field used to align the nuclei).

Research by Damadian has shown that the T1 and T2 were longer for hydrogen nuclei in diseased tissue, like cancer, than in normal tissue. And the patent at issue in this case specifies use of T1 and T2 to discriminate between diseased and normal tissue.

But in reversing the jury verdict of patent infringement, U.S. District Court Judge Robert E. Keeton said that an MRI device "need not, and does not as routinely used, compute T1 and T2 values." Moreover, he wrote, though T1 and T2 are by far the greatest contributors to MRI signal strength, "neither that particular value [of signal strength], nor the value of any component that contributes to it, is itself significant to imaging." In fact, he said, detecting cancer with an MRI machine depends not on T1 and T2 "but instead upon inferences of a diagnostician from the shapes and locations of images, and from the diagnostician's comparison of images rather than comparison of T1 and T2 values." Therefore, he said, Technicare's MRI machines do not infringe on Damadian's patent.

According to Damadian, "That interpretation by the judge is nonsense," because "when you make an [NMR] image, the image is of T1 and T2." So the difference in those values for normal and abnormal tissue is integral to a diagnostician's use of any commercial MRI system, Damadian maintains -- "and there was never any testimony [in the trial] contrary to that."

"We feel confident that the judge's decision will be reversed and the jury verdict reinstated," says John Nelson, an attorney representing Fonar and Damadian. But even winning the appeal would not settle the issue: Deciding what "reasonable royalties" the patent infringer owed could take another year.