Fonar Corp. v. General Electric Co.

902 F. Supp. 330, 41 U.S.P.Q. 2d (BNA) 1088, 1995 U.S. Dist. LEXIS 14931, 1995 WL 603524

• Court: District Court, E.D. New York
• Decided: September 30, 1995
• Docket: CV 92-4196
• Status: Published

Opinion

WEXLER, District Judge.

Fonar Corporation and its founder Dr. Raymond V. Damadian (“Damadian”) (collectively, “Fonar”) brought the above-referenced patent infringement action, by summons and complaint dated September 1, 1992, against General Electric Corporation and Drueker & Genuth, MDs, P.C., d/b/a South Shore Imaging Associates (“South Shore”) (collectively, “GE”). 1 On May 19, 1995, after a thirteen-day trial on liability, a jury returned a verdict in favor of plaintiffs wherein it found that the two patents in suit were not invalid and infringed. On May 26, 1995, after hearing three days of testimony on damages, the same jury awarded plaintiffs $110,575,000. Presently before the Court is GE’s motion, pursuant to Fed.R.Civ.P. 50 and 59, for judgment as a matter of law (“JMOL”) and, in the alternative, for a new trial.

I. THE ’882 PATENT

A. Background

Fonar Corporation is the exclusive licensee of U.S. Patent No. 3,789,832 (the “’832 patent”), entitled “Apparatus and Method for Detecting Cancer in Tissue,” issued February 5, 1974 to Damadian, on application filed March 17, 1972. The method by which cancer may be detected, as claimed by the ’832 patent, employs the use of nuclear magnetic resonance (“NMR” or “MR”). Some technological background is necessary. 2

Nuclei in the body’s atoms generate magnetic fields at varying energy levels. Under normal conditions, the north and south poles of the body’s nuclei point in various directions. When placed in an external magnetic field, however, nuclei align with or against the direction of the external field depending on the degree of their own magnetic strength — nuclei in a low-energy state align with the external field; those in a high-energy state point away from the field. Exposure to an external magnetic field also causes nuclei to wobble, or precess, as they spin on their own axes.

■ A radio frequency (“RF”) wave, when pulsed into a body exposed to an external magnetic field, produces two effects in nuclei: first, the RF energy causes low-energy nuclei to flip into a high-energy state and align in a direction opposite to the magnetic field; second, it causes nuclei to precess in synch, or “in phase.” When the RF pulse is eliminated, the effected nuclei return to their natural magnetic state. This is called “relaxation.” Relaxation is quick, but not instantaneous.

During relaxation, effected nuclei emit a detectable RF signal of their own as they return to their low energy state. Their signal comprises three elements. First, the strength of the signal depends on the number of nuclei that flipped into high energy state when the RF pulse was turned on. This element is called proton density. Second, the strength of the signal depends on the rate at which the effected nuclei return to their low energy state, or, to use the proper term, the rate at which they achieve “spin-lattice” relaxation. This time is specified by “Tj”. The last element effecting the strength of the emitted signal is the rate at which the effected nuclei loose “phase” among their respective wobbles, or, achieve “spin-spin” relaxation. This time is specified by “T2”.

Trial testimony revealed that, in 1970, Da-madian conducted research to test his theory that Ti and T2 relaxation times were different, in healthy and diseased tissue of the same tissue type. After obtaining encouraging results, Damadian applied for and obtained the ’832 patent. Claims 1 and 2 of the ’832 patent read:

1. A method for detecting cancer comprising:

a. measuring and establishing standard NMR spin-lattice relaxation times and spin-spin relaxation times for both normal and cancerous tissue of the type under analysis using as an indicator nuclei at least one nuclei which exhibits deviant behavior in cancerous tissue;

b. measuring the NMR spin-lattice relaxation times and spin-spin relaxation times for the suspected tissue to determine the extent of deviant behavior of the indicator nuclei; and

c. comparing the values obtained in (b) against the standards obtained in (a).

2. The method of claim 1, wherein the indicator nuclei are cell water protons.

In the instant action, Fonar alleged that, between September 1986 and February 3, 1991, GE infringed or induced infringement of Claim 1 and dependent Claim 2 of the ’832 patent by the manufacture, use, and/or sale of magnetic resonance imaging (“MRI”) scanners.

MRI scanners read the RF signal emitted from nuclei during relaxation and produce an image therefrom. The image represents a view of a shoe of the body at a particular point along a particular plane. The scanner selects the slice of the body to be imaged and subdivides it into tiny squares of tissue, or voxels, through the use of gradient magnetic fields. After the RF pulse is introduced, the scanner’s antenna collects an RF signal emitted from each voxel during relaxation. The electric signals are then converted into values to be read by a computer. For each voxel there is a corresponding pixel on an imaging screen. The computer displays the value, or RF signal strength, for each voxel as a shade of gray on the corresponding pixel. How dark or light a pixel is depends on the strength or weakness of the RF signal emitted by its corresponding voxel.

The scanner cannot produce a useful image, however, without some assistance from the operating technician. Nuclei of a given tissue type generally emit RF signals of the same strength during relaxation. As such, that tissue type should appear in the MR image as a single shade of gray. An abnormality in that tissue, however, will produce an RF signal of a different strength and a different shade of gray on the screen. As such, a dark spot on an otherwise light image of brain tissue, for example, would indicate to a physician that there is an abnormality in the patient’s brain. This being so, it is essential that the scanner obtain the maximum contrast between the varying degrees of gray. The scanner accomplishes this by repeating the gradient-RF pulse process many times. If the pulsing is done with the proper timing, the contrast will increase and the image will become sharper with every repetition. The technician sets the timing, or, pulse sequence, as it is called. This task actually involves two settings: time echo (“TE”) and time to repeat (“TR”). TE and TR are called pulse protocols or timing parameters.

Fonar contended that GE infringed the ’832 patent because MRI scanners manufactured and sold by GE determine whether cancer may be present in suspect tissue by comparing the Ti or T2 relaxation times of the suspect tissue with predetermined standard Ti or T2 relaxation times for normal and cancerous tissue of the same tissue type. The jury agreed, finding that GE infringed Claims 1 and 2 of the ’832 patent. As to validity, the jury found that “Damadian invented] the subject matter claimed by the [’832] patent.” The jury awarded Fonar $35 million in damages. The Court reserved decision on GE’s post-trial motion for JMOL. The instant motion for JMOL is made on renewal.

B. Claim Construction and Infringement

On a motion for JMOL, a district court may not disregard a jury verdict unless it is convinced that the trial record contains no evidence “ ‘upon which a jury might properly have returned a verdict in [the non-movant’s] favor when the correct legal standard is applied.’” Markman v. Westview Instruments, Inc., 52 F.3d 967, 975 (Fed.Cir.1995) (en banc) (quoting Jamesbury Corp. v. Litton Indus. Prods., Inc., 756 F.2d 1556, 1560 (Fed.Cir.1985)) (emphasis omitted). “Factual findings made by the jury in arriv ing at its verdict are to be upheld unless the party moving for JMOL shows that (when the correct legal standard is applied) there is not substantial evidence to support a finding in favor of the nonmovant.” Id. (citing Read Corp. v. Portec, Inc., 970 F.2d 816, 821 (Fed. Cir.1992)). Whether the correct legal standards have been applied by the jury, either expressly or by implication, is a question considered by the district court de novo. Id. (citations omitted).

“Substantial” evidence is such relevant evidence, taken from the record as a whole, as might be accepted by a reasonable mind as adequate to support the finding under review. Dana Corp. v. IPC Ltd. Partnership, 860 F.2d 415, 417 (Fed.Cir.1988), cert. denied, 490 U.S. 1067, 109 S.Ct. 2068, 104 L.Ed.2d 633 (1989). In determining whether substantial evidence is present, “a court must: (1) consider all the evidence, (2) in a light most favorable to the non-mover; (3) drawing all reasonable inferences favorable to the non-mover; (4) without determining credibility of witnesses, and (5) without substituting its choice for that of the jury between conflicting elements in the evidence.” Id. (quoting Connell v. Sears, Roebuck & Co., 722 F.2d 1542, 1546 (Fed.Cir.1983)) (other citations omitted).

A jury’s infringement verdict must stand where a court finds substantial evidence supporting the finding that the accused process or product performs each element of at least one of the properly construed patent claims. See Read Corp. v. Portec, Inc., 970 F.2d 816, 821 (Fed.Cir.1992). The analysis comprises two determinations: the court must define, as a matter of law, the meaning and scope of language used in the patent claims, and must also decide whether the jury’s infringement verdict is supported by substantial evidence in light of the properly construed claims. Markman, 52 F.3d at 976, 979. 3

It is well settled that “ ‘[a] patent covers the invention or inventions which the court, in construing its provisions, .decides that it describes and claims.’” Id. at 979 (quoting 3 William C. Robinson, A Treatise on the Law of Patents for Useful Inventions § 1019, at 247 (1890)). In defining the meaning and scope of patent claims, a district court must consider: the language of the claims; the patent’s specification; and the patent’s prosecution history. Id.; Unique Concepts, Inc. v. Brown, 939 F.2d 1558, 1561 (Fed.Cir.1991). A court in its discretion may also consider extrinsic evidence, such as expert and inventor testimony, dictionaries, and learned treatises. Markman, 52 F.3d at 980. It is the language of the claims, however, that defines the scope of the invention; other evidence should be considered only insofar as it informs the court as to the meaning of those words. The words of the claims should be given their ordinary meaning; for a word to take on some special meaning, it must be defined as such in the specification. See id. at 979-80 (citing Intellicall, Inc. v. Phonometrics, Inc., 952 F.2d 1384, 1388 (Fed.Cir.1992)). In this sense, the specification serves as “a sort of dictionary, which explains the invention.” Id.

Element (a) of Claim 1 of the ’832 patent identifies the first step for cancer detection as the “measuring and establishing [of] standard NMR [Tj] relaxation times and [T2] relaxation times for both normal and cancerous tissue.” Element (a) claims not merely the measurement and establishment of any data for Ti and T2, but limits the invention to the measurement and establishment of data suitable for serving as a “standard.”' Given the comparison required by element (c), the word “standard” in (a) is a significant limitation. Having read the claims, specification, and prosecution history of the ’832 patent, and having heard the testimony of Damadian, the inventor, and Reuben Mezrich, GE’s claim interpretation expert, the Court interprets “standard” to mean NMR relaxation time data against which other data may be compared in order to detect cancer. Element (b) defines what the “other” data must be. Element (c) requires that the detection of cancer be performed by a comparison of “the values obtained in (b) against the standards obtained in (a).” As the specification aptly summarizes, at column 2, lines 22-26, Claim 1 of the ’832 patent disclose a process for determining whether cancer is present “[b]y comparing the relaxation times for [suspect] tissue with relaxation times obtained for known normal tissue and known cancerous tissue.”

GE’s principal argument on its motion for JMOL is that the properly construed claims of the ’832 patent do not read on the process of imaging. GE contends that its scanners do not infringe the ’832 patent because they merely produce images that can be analyzed by physicians in order to determine whether internal tissues contain abnormalities. If cancer can be detected, GE argues, the detection is performed by the physician, not the scanner. Even under a broad reading of the claims, GE insists, the ’832 patent does not contemplate the process of imaging. Indeed, the words “image” or “imaging” appear nowhere in the claim language or specification.

Fonar would agree that, under GE’s understanding of the process, the claims of the ’832 patent do not read on imaging. But Fonar offers a different explanation for how cancer is detected by MRI scanners made and sold by GE. Under Fonar’s theory, the scanner does the detection, and it does so by comparing the relaxation times of suspect tissue against standard relaxation times for normal and cancerous tissue. This process, Fonar argues, comes squarely within the scope of the properly construed claims.

In order to establish that GE performed element (a) of the ’832 patent, Fonar had to prove that GE measured and established standards—that is, Tj and T2 relaxation time data for normal and cancerous tissue against which other data could be compared in order to achieve cancer detection. Fonar’s theory, as presented mainly through the testimony of Damadian, is as follows. GE’s scanners are able to detect cancer in a given tissue type only when they are properly calibrated for that tissue type. Proper calibration is achieved only by using an optimal pulse sequence—that is, the pulse protocol settings (TE and TR) that provide maximum differential in pixel brightness. Proper TE and TR settings for a given tissue type can be determined only by way of mathematical formulae, within which the only variables are “standard” Ti and T2 relaxation times for normal and cancerous tissue. The standards—numerical values—vary depending on the type of tissue under examination. In sum, Fonar contends that a useful image cannot be obtained without an optimal pulse sequence, and that an optimal pulse sequence cannot be obtained without knowing the “standard” values for that tissue type.

Fonar identifies several documents in evidence in support of the proposition that GE measured and established standards, and, thereby, performed step (a) of Claim 1. The first is the back cover of a scientific journal, which contained an article written by GE’s Manager of NMR Applications, Felix W. Wehrli. The article contains the following language: “Once detailed knowledge of Ti and T2 data exists for various tissue types (primarily normal and diseased), pathology-specific scan protocols can be established. This is one goal of General Electric’s extensive applications development program.” PI. Exh. 612. The second is a 1987 article de tailing the results of five years of GE research. Authored by GE scientists Paul A. Bottomley (“Bottomley”) et al., the article includes a table of “relaxation times for different pathologies and their dependence on NMR frequency, tissue of origin, temperature, and species.” PI. Exh. 599, at 2. The article states: “Such information can be used for optimizing image contrast between normal and pathological tissue via appropriate adjustment of pulse sequence timing parameters, and may be of diagnostic value.” Id. The article also referred to a precursor work, co-authored by Bottomley in 1984, which reviewed “normal tissue relaxation.” Id. at 1. Fonar contends, based on this evidence, that GE measured and established standards.

The evidence strongly supports the proposition that Bottomley’s findings constitute essential ingredients in the process of optimizing pulse sequences. The evidence also suggests that optimal TE and TR settings are essential to the production of a useful MR image, an image that undeniably helps physicians determine whether cancer is present in tissue. But, in order for Bottomley’s figures to constitute “standards,” as the Court has construed that term, his data must be suitable for comparison against other data so as to form a basis for cancer detection. Comparisons occur during the course of an MR scan. In a T^weighted image of brain tissue, for example, there is the comparison on the screen between light gray, indicating normal tissue, and dark gray, indicating diseased tissue. In a sense, the scanner’s computer compares the RF signal it receives from a particular voxel against its internal programming in order to determine the signal’s strength and set the appropriate brightness for the corresponding pixel. It cannot be said, however, that the scanner compares Ti and T2 relaxation times for the tissue under examination against “standard” Tj and T2 relaxation times for normal and cancerous tissue in order to produce an image useful for cancer detection. The image is a reflection, nearly a direct reflection, of Ti and T2 relaxation times for the tissue under examination. It is an indication, not a comparison. Bot-tomley’s research provides for a more precise indication; it does not, however, provide the basis for a comparison.

' A similar, but not the same, conclusion was reached several years ago in Fonar Corp. v. Johnson & Johnson, 630 F.Supp. 581 (D.Mass.1986), aff'd, 821 F.2d 627 (Fed.Cir.1987), ce rt. denied, 484 U.S. 1027, 108 S.Ct. 751, 98 L.Ed.2d 764 (1988). In Johnson & Johnson, a jury verdict of infringement of Claims 1 and 2 of the ’832 patent was disregarded by the district court on the defendants’ motion for JMOL because “the revealing comparison [was] not between Ti and T2 values for suspected tissue of a particular type ..., on the one hand, and standard values, on the other hand.” Id. at 585. The district court found Fonar’s argument that the requisite comparison existed between imaged values presented on a screen and “standard” imaged values stored in a physician’s memory to be “wholly beyond the scope of any permissible interpretation.” Id. at 586-87. The district court noted, however, that although the evidence had not supported the proposition that “accumulated scientific knowledge about Tj and T2 values for various types of tissue [was] sufficient ... to enable anyone ... to prepare a set of standards” as contemplated by the ’832 patent, “[a]n optimist may hope that it will yet be possible.” Id. at 587-88. Fonar’s theory in the instant case, therefore, was that Bottomley and GE had turned that possibility into a reality.

Because its theory of infringement was different, the evidence presented by Fonar in the instant case differed significantly from that presented in Johnson & Johnson. Although tempting, it would have been improper under these circumstances for the Court to grant GE’s pre-trial motions on collateral estoppel grounds. In retrospect, however, it is fair to categorize Fonar’s argument to the Court in the instant case as a valiant effort to fit a different square peg into the same round hole. Again, it failed to establish the existence of standard Ti and T2 values as contemplated by Damadian. As such, the Court determines, as a matter of law, that the infringement verdict returned by the jury is not supported by substantial evidence. GE’s scanners do not detect cancer by the comparison method disclosed by Claims 1 and 2 of the ’832 patent. The motion for JMOL is granted on the ground that GE’s scanners do not infringe Claims 1 and 2 of the ’832 patent as construed by the Court. In doing so, the Court must disregard the jury’s answers to question I.A. on the Verdict Sheet. 4

C. Invalidity

Pursuant to 35 U.S.C. § 282, a patent is presumed valid. A party challenging this presumption, must prove by clear and convincing evidence that the patent is invalid. North Am. Vaccine v. American Cyanamid Co., 7 F.3d 1571, 1579 (Fed.Cir.1993), ce rt. denied, — U.S. -, 114 S.Ct. 1645, 128 L.Ed.2d 365 (1994). On motion for JMOL, the Court analyzes invalidity issues submitted to the jury by the same standards used for infringement—that is, (1) whether there is substantial evidence supporting the jury’s express or implied findings of fact and (2) whether those findings are legally sufficient to support the legal conclusion drawn by the jury in reaching its verdict. See Dana Corp., 860 F.2d at 417.

The Court concludes that there is substantial evidence in the record supporting the jury’s finding that the ’832 patent was not procured through inequitable conduct. In addition, the Court determines that, based on the evidence presented by GE, a reasonable jury could find no set of facts upon which an obviousness determination could be based. As such, GE failed to sustain its burden of proving, by clear and convincing evidence, that Claims 1 and 2 of the ’832 are invalid. Accordingly, the motion for JMOL is denied in this respect.

II. THE ’966 PATENT

Fonar Corporation is the exclusive licensee of U.S. Patent No. 4,871,966 (the “’966 patent”), entitled “Apparatus and Method for Multiple Angle Oblique Magnetic Resonance Imaging,” issued October 3, 1989 to Stanton D. Smith (“Smith”), David Hertz (“Hertz”), Robert Wolf (“Wolf’), and Robert H. Olsen (“Olsen”), on application filed November 16, 1988. The ’966 patent discloses a method and apparatus for obtaining, in the course of a single MR scan, a number of images at different angles to each other.

The imaging technique claimed by the ’966 patent, called multi-angle oblique (“MAO”), is particularly useful for scanning the spine. In order to analyze spinal tissue, a physician needs images representing an axial view of each disc—that is, a picture of what each disc would look like if it were cut in half along its east-west axis. Because vertebrae join each other at a variety of angles to constitute the snake-like spine, the proper angle from which to produce an image of one disc will differ from the next. In the past, each disc was scanned separately, resetting the scanner for the appropriate angle each time. Using MAO, the scanner can produce images of all the discs, at desired angles, in one scan. MAO, as claimed by the ’966 patent, involves the use of a “scout image.” The scout image is a preliminary sagittal-slice image of the spine—that is, what the spine would look like if it were cut in half along its north-south axis. By placing cursor lines across the scout image, at each disc, the technician selects the angle from which each disc will be imaged. With the push of a button, the scanner generates axial pictures of each disc and a scout picture identifying the angles from which each axial picture was taken.

In the instant action, Fonar alleged that GE infringes or induces infringement of Claims 1, 2, 4, 5, and 12 of the ’966 patent by the manufacture, use, sale, and maintenance of MRI scanners. At trial, GE defended on the following grounds: (1) that the ’966 patent was invalid, pursuant to 35 U.S.C. § 112, ¶ 1, for failure to disclose “the best mode contemplated by the inventorfs] of carrying out [their] invention” and information sufficient to enable one skilled in the art to make and use the claimed invention without the necessity of undue experiment; (2) that the ’966 patent was invalid, pursuant to 35 U.S.C. § 102(f), for failure to disclose all the inventors; (3) that the ’966 patent was invalid, pursuant to 36 U.S.C. § 102(b), because its subject matter was in public use more than one year prior to the filing date; (4) that the ’966 patent was invalid, pursuant to 35 U.S.C. § 103, because its subject matter would have been obvious to one skilled in the prior art; (5) that the ’966 patent was invalid, pursuant to 35 U.S.C. § 102(g), because its subject matter was anticipated by the prior invention of another; and (6) that GE did not infringe or induce infringement of the ’966 patent.

The jury found that GE infringed Claims 1, 2, 4, 5, and 12 of the ’966 patent, and, in response to interrogatories submitted, pursuant to Fed.R.Civ.P. 49(a), on each invalidity defense, that the claims were not invalid. As to damages, the jury awarded Fonar $27,-825,000 in lost profits for GE’s sales of 75 MRI scanners, and $34,125,000 in reasonable royalties for GE’s sales of 525 scanners. The jury also awarded Fonar $13,625,000 in damages for inducement of infringement. This last award was given in response to an interrogatory, which the Court gave based on Fonar’s theory that GE’s performance of maintenance and repair after notice of infringement on scanners sold prior to notice of infringement constituted inducement of infringement.

GE moved for JMOL on numerous grounds at the close of the proceedings. The Court denied the motions on all grounds except failure to prove inducement of infringement, a ground upon which the Court reserved decision. GE has renewed its motion on every ground—twenty-three to be exact. The motion is summarily denied as to all grounds not specifically addressed by the Court in this Memorandum & Order.

B. Claim Construction

As stated above, it is the obligation of the Court to render an interpretation of the claims of the patent-in-suit. The full text of the claims at issue are set forth in the Appendix to this Memorandum & Order.

Claim 4 and dependent Claim 5 contain “means-plus-function” language. Claims with means-plus-function language “cover the corresponding structure, material, or acts described in the specification and equivalents thereof.” 35 U.S.C. § 112; King Instruments Corp. v. Perego, 65 F.3d 941, 945 (Fed.Cir.1995). This rule of construction applies whether the claim language is being construed in the context of an infringement or invalidity determination. In re Donaldson, 16 F.3d 1189, 1193 (Fed.Cir.1994).

Element (a) of Claim 4 covers: “positioning an object in an NMR imaging apparatus which includes ... means for producing an image from said NMR imaging data.” Element (b) covers: “operating said NMR imaging apparatus to obtain NMR scout image for a portion of said object.” Element (c) covers: “using said scout image to select a first plane and a second plane of said object for which NMR imaging data is to be obtained.” Read in conjunction with its preceding elements, element (c) limits the invention in the following respect: each of the multiple angles at which images are to be obtained must be “selectfed]” by “using” a “scout image.”

The specification of the '966 patent makes two references to the phrase “scout image.” It states in column 3, lines 29-35, that a scout image is an image of a “portion of the object of the examination.... used to select” planes from which “NMR image data is to be obtained.” It also states, in column 16, lines 53-57:

Referring to FIG. 10, a particular medical use of the apparatus of the present invention is depicted. A patient is disposed in an NMR imaging apparatus, and a scout scan is taken of the patient’s spine, which is displayed on a screen 130.

The Court construes the phrase “scout image” to mean a preliminary NMR image of a portion of the object of the examination used for the purpose of “selecting]” desired imaging planes. In so construing, the Court declines to limit the phrase “scout image” to preliminary images of the spine. See Electro Med. Sys. S.A. v. Cooper Life Sciences, 34 F.3d 1048, 1054 (Fed.Cir.1994) (cautioning that “although the specification may well indicate that certain embodiments are preferred, particular embodiments appearing in the specification will not be read into the claims when the claim language is broader than such embodiments”).

The limitation also requires that the desired imaging planes be “select[ed],” by “using” the scout image. These words identify the method by which the imaging angles must be selected. Figures 5 and 10, and corresponding text in the specification, show quite clearly that the manual use of cursor lines on the scout image are the method for selecting imaging planes. The specification does indicate that “direct input” of desired imaging planes “suffices for the purposes of the present invention,” see ’966 patent, col. 15, lines 39-42, but because Claim 4 is limited by its own language to use of a “scout image,” direct input of desired imaging planes is not covered by Claim 4. See In re Donaldson, 16 F.3d at 1195. The Court determines that Claim 4 covers the manual use of cursor lines on a scout image to select desired imaging angles.

Claim 1 and dependent Claim 2, however, are not limited to use of a “scout image.” As such, the Court determines that direct input of desired imaging planes is covered by Claim 1.

As indicated above, a party challenging the presumption that a patent is valid, must prove by clear and convincing evidence that the patent is invalid. North Am. Vaccine, 7 F.3d at 1579. Based on its responses to interrogatories, the jury found that GE failed to prove that the ’966 patent was invalid. GE moves for JMOL on several grounds. The verdict must stand, however, unless GE establishes that the jury’s express or implied findings of fact are not supported by substantial evidence or that those findings are insufficient to support the legal conclusion drawn by the jury in reaching its verdict. See Dana Corp., 860 F.2d at 417.

1. Best Mode

GE argues that the ’966 patent is invalid because its inventors contemplated a best mode for carrying out Claims 1, 2, 4, 5, and 12 which they did not adequately disclose. The first paragraph of 35 U.S.C. § 112 sets forth the requirement that the patent specification “set forth the best mode contemplated by the inventor of carrying out his invention.” Best mode analysis has two components. In re Hayes Microcomputer Prods. Patent Litig., 982 F.2d 1527, 1536 (Fed.Cir.1992) (citing Chemcast Corp. v. Arco Indus. Corp., 913 F.2d 923, 927 (Fed.Cir.1990)).

The first inquiry focuses on whether the inventor knew of a mode of practicing his invention at the time he filed his patent application which he considered to be better than any other. This determination is subjective, focusing on the inventor’s state of mind at the time he filed his application. If he did have a best mode, the next question is whether he disclosed it and .did so adequately to enable one of ordinary skill in the art to practice the best mode. This is an objective determination. There must be no concealment of a mode known by the inventor to be better than that which is disclosed.

Id. (citing Hybritech Inc. v. Monoclonal Antibodies, Inc., 802 F.2d 1367, 1384-85 (Fed.Cir.1986), cer t. denied, 480 U.S. 947, 107 S.Ct. 1606, 94 L.Ed.2d 792 (1987)). Whether the best mode requirement has been met is a question of fact. Id.

GE asserts that, at the time the inventors filed the patent application, the inventors knew that certain computer hardware and software were better than any other for implementing MAO, as claimed by the ’966 patent. GE argues that the inventors’ bare functional disclosure in the ’966 patent effectively concealed their best mode. Fonar acknowledges that the inventors contemplated particular hardware and software as part of their best mode for implementing MAO, but contends that the jury verdict is supported by substantial evidence because this mode was adequately disclosed in the ’966 patent. 5

In discussing the best mode requirement in a ease where the patent-in-suit did not disclose a specific computer program used to implement a system for converting a seismic time section, the United States Court of Customs and Patent Appeals stated:

In general, writing a computer program may be a task requiring the most sublime of the inventive faculty or it may require the droning use of clerical skill. The difference between the two extremes lies in the creation of mathematical methodology to bridge the gap between the information one starts with (the “input”) and the information that is desired (the “output”)....

In re Sherwood, 613 F.2d 809, 817 (C.C.P.A.1980), ce rt. denied, 450 U.S. 994, 101 S.Ct. 1694, 68 L.Ed.2d 193 (1981). Continuing, the court in Sherwood held:

[Because the] specification provides the general mathematical equations used and teaches the further “trick” of chopping the physical input seismic traces into segments via a mathematical manipulation,.... [it] delineates the best mode in a manner sufficient to require only the application of routine skill to produce a workable digital computer program.

Id. In fact, disclosure of merely the general function of hardware and software necessary for implementation of an invention can satisfy the best mode requirement. See In re Hayes, 982 F.2d at 1537.

In order to perform MAO—that is, to generate multiple images at different orientations during the course of a single scan—an MRI scanner must have the capacity, among other things, to simultaneously launch two gradient waveforms towards each other from transverse angles. It must also be able to assign each gradient a different frequency. When the gradients collide, a new, hybrid gradient is created along a new, oblique-angle plane. The strength chosen for each of the initial gradients determines the exact angle of the hybrid gradient. As Wolf, one of the inventors on the ’966 patent, explained at trial, it is “like turning on the hot and cold water at the same time to get the warm water, [only] now we’re mixing the two power supplies.” Trial Tr. at 484. This process is called rotating gradients.

The inventors on the ’966 patent believed that gradient rotation could best be accomplished with hardware that they called a “gradient multiplier board” (“GMB”). The GMB was engineered by Hertz. On his direct examination, the following colloquy took place between Hertz and Fonar’s counsel:

Q. Now was this board disclosed in the [’966] patent?

A. Yes, it was.

Q. And is it called the gradient multiplier board on Fonar (sic)?

A. That’s correct.

Q. And would you show the ladies and gentlemen of the jury where ... on figure seven [of the ’966 patent], there is a blocked diagram of the gradient multiplier board?

A. The area on the dotted line.

Q. And in the—within that dotted line and in the description of the patent does it tell the person of ordinary skill in the art everything they need to know in order to build a gradient multiplier board that will operate on a general MRI system?

A. Yes, it does. It is black dark art (sic) that will include all the functions that will be needed.

Q. And is there then a written description ... in the body of the patent that goes along with those functions?

A. Yes, there is.

Trial Tr. at 640-41.

The specification discloses that the GMB comprises six principal components—an arithmetic unit, a multiplier and offset parameter RAM, a slice pointer, a level pointer, and two terminals, see ’966 Patent, fig. 7— and describes how they interact during the course of a MAO scan. For example, it states that: “[t]he slice pointer 31, in response to a pulse from the pulse programmer 24, outputs a signal to the RAM 28,” id. at col. 13, lines 1-3; “[t]he waveform data and the axis signal are conveyed to the arithmetic unit 25, and the axis signal is conveyed to the multiplier and offset parameter RAM 28,” id. at col. 13, lines 25-28; “the RAM 28 outputs the above-described multiplier and offset terms,” id. at col. 13, lines 35-36. More specifically, the specification describes how the GMB generates a gradient waveform at the desired strength:

[fjor an input from the slice pointer 31 indicating the second slice, and an input from the generator 20 indicating the Z axis, the RAM 28 outputs the multiplier [COS(d) + SIN(d) ]CSS, and outputs the offset A[COS(d) - SIN(d) ]CSS. The arithmetic unit 25 multiplies the digital data representing the generic waveform G(t) by the multiplier term and to this expression adds the offset term. The sum, which represents the waveform segment to be applied to the Z coil, is output in digital form from terminal 41, and a signal indicating the corresponding Z axis is output from the terminal 32.

Id. at col. 13, lines 37-47.

The testimony of Smith, the lead inventor on the ’966 patent, supports the sufficiency of the disclosure. After identifying one of the “benefits” of the GMB as its capacity to allow “the scanner to go through the entire data acquisition process without ever talking to the host computer for that purpose,” the following colloquy took place between Smith and counsel for GE:

Q. Does [the ’966 patent] tell you how to do it?

A. I think it shows essentially how these parameters are stored in the multiplier and offset parameter RAM, and it shows there is an arithmetic you want which does the calculations and there is some reference to that whole process there.

There [are] some equations that show, for example, how two gradients may be combined arithmetically. That particular equation describes [the] kind of calculation that leads to rotation of a plane.

... I think I would say yes to your question.

Trial Tr. at 1264-66. Smith’s testimony is particularly probative given the fact that he was the only inventor not employed by Fonar at the time of trial, and that he was called as GE’s witness. That and other evidence supports the conclusion that a person of ordinary skill in the relevant art could engineer a GMB based on the information provided in the ’966 patent. See In re Hayes, 982 F.2d at 1537-39.

GE, nonetheless, argues that the best mode was not disclosed because the ’966 patent makes no reference to the existence or appropriate function of “comparators,” which, according to Hertz, exist on the GMB for interaction with power supplies. Indeed, Hertz testified on cross-examination that the comparators were “very important”; he referred to them, in fact, as the “trick” of the invention. See Trial Tr. at 677-78. On redirect, however, he testified as follows:

Q. Are [comparators] necessary on every MRI machine?

A. No, they are not.

Q. If someone is using an MRI machine that requires a comparator, will they be able to find out that they have to put a comparator in when they build the multi-angle oblique invention from reading the patent?

A. I believe that a skilled engineer will have the ability to understand it.

Id. at 689. Even if the cross-examination testimony of Hertz establishes that the inventors contemplated the use of comparators as part of the best mode of practicing their invention, his testimony on redirect makes clear that best mode was adequately dis-' closed without reference to comparators. As such, the Court determines that substantial evidence supports the jury’s implied finding that the GMB board was adequately disclosed in the ’966 patent.

GE also contends that the inventors failed to adequately disclose information necessary to write the software for the MAO invention. MAO requires two types of software. One type, which is generally referred to as a “device driver,” allows the host computer to exercise and operate hardware that is exter nal to the computer — that is, the GMB. Wolf, after reviewing a stack of GMB specifications given to him by Hertz, developed two device driver subroutines. The other type of software, written by Olsen, allows the scanner technician to choose imaging slices on the scout image. Based on the chosen slices, his program calculates and communicates the appropriate parameters to Wolfs device drivers, which, in turn, run the GMB. This type of software is called a “user interface” or “graphic prescription.” GE contends that the inventors failed to meet the best mode requirement because the ’966 patent disclosed neither the specifications for the software itself nor the GMB specifications necessary to write the device driver subroutines.

The flaw in GE’s argument is apparent from the following testimony given by Wolf:

Q. From that written description [of the GMB in the ’966 patent], is there sufficient description to a software engineer, such as yourself, of what software needs to be written in order to perform the multi-angle oblique invention?

A. Yes.

Q. In any event, the software, itself, as we see in the hundred pages of Exhibit 816, is not reproduced in the patent.

Is that right?

A. That is correct.

Q. Why is that?

A. For a few reasons.

First of all, it’s large....

What’s much more important is to have a description of what the software has to do, and that is what you will find in the patent.

Id. at 508-10. As an example, Wolf explained that the specification teaches “how the computer can calculate the various frequencies for each slice for each angle based on the strength of the slice of the gradient when its rotated to the angle of the slice.” Id. at 628 (referring to ’966 patent, col. 12, lines 16-23 6 ). The language identified, according to Wolfs testimony, “tells someone what to base the calculations on for the RF frequencies.” Id. at 629.

The evidence supports the conclusion that a software engineer of ordinary skill could design software sufficient to run hardware capable of performing the invention claimed by the ’966 patent. See In re Hayes, 982 F.2d at 1537-39.

The Court determines that there is substantial evidence supporting the jury’s finding that GE failed to prove, by clear and convincing evidence, that the best mode requirement was not violated. 7

2. Enablement

GE also contends that the Claims 1, 2, 4, 5, and 12 are invalid because the disclosure in the ’966 patent was insufficient to enable one skilled in the art to make and use the invention. The first paragraph of 35 U.S.C. § 112 contains a second disclosure requirement that is separate and distinct from best mode. It mandates that the specification contain a written description of the claimed invention sufficient to enable any person skilled in the art to which the invention pertains to make and use that invention. Although not explicitly stated in § 112, to be enabling, the specification must teach those skilled in the art how to make and use the full scope of the claimed invention without “undue experimentation.” In re Wright, 999 F.2d 1557, 1561 (Fed.Cir.1993). Nothing more than an objective enablement is required, however; it is irrelevant whether the teaching is provided through broad terminology or illustrative examples. Id. (citing In re Marzocchi, 439 F.2d 220, 223 (C.C.P.A.1971)).

Enablement is a question of law, supported by underlying factual determinations. See Quaker City Gear Works, Inc. v. Skil Corp., 747 F.2d 1446, 1453-54 (Fed.Cir.1984), ce rt. denied, 471 U.S. 1136, 105 S.Ct. 2676, 86 L.Ed.2d 694 (1985). A court may submit legal issues such as enablement to the jury, pursuant to Fed.R.Civ.P. 49(a), but if it does, “the court may not make subsequent findings which overrule an implicit and inherent finding of the jury.” Id.

The jury responded in the affirmative to an interrogatory asking whether the description in the ’966 patent was sufficient to enable one skilled in the art to make and use the subject matter of Claims 1, 2, 4, 5, and 12. Implicit therein is the jury’s factual determination that skilled engineers could make and use the full scope of the invention without unreasonable experimentation. GE asserts that, given the disclosure in the ’966 patent, a skilled engineer would not be able to build the MAO invention in less than one to three years. GE’s assertion, however, is supported by no more than its witnesses’ conclusory oral testimony that the hardware and software specifications were insufficiently disclosed. As discussed above, ample evidence supports the conclusion that the inventors disclosure was adequate in this regard.

The record supports the jury’s implied conclusion that one skilled in the art could build the invention without undue experimentation. The Court determines, as a matter of law, that the description of the invention in the ’966 patent is enabling.

3. Anticipation: The Work of Hattes

A patent may be invalid as anticipated pursuant to 35 U.S.C. § 102(g) where there is prior conception and reduction to practice by another of the patentee’s invention. See Texas Instruments, Inc. v. United States Int’l Trade Comm’n, 988 F.2d 1165, 1177 (Fed.Cir.1993). Although anticipation is a question of fact, conception and reduction to practice are questions of law. See Hybritech Inc. v. Monoclonal Antibodies, Inc., 802 F.2d 1367, 1376, 1379 (Fed.Cir.1986), cert. denied, 480 U.S. 947, 107 S.Ct. 1606, 94 L.Ed.2d 792 (1987).

GE argues that its scientists were first to invent the subject matter of Claims 1, 2, 4, 5, and 12 of the ’966 patent. “Priority goes to the first party to reduce the invention to practice unless the other party can show that it was the first to conceive the invention and that it exercised reasonable diligence in later reducing that invention to practice.” Price v. Symsek, 988 F.2d 1187, 1190 (Fed.Cir.1993). Conception is the “ ‘formulation in the mind of the inventor, of a definite and permanent idea of the complete and operative invention, as it is hereafter to be applied in practice.’ ” Id. (quoting 1 Robinson on Patents 532 (1890)). Reduction to practice requires that the claimed invention work for its intended purpose. Id. It is undisputed that the inventors on the ’966 patent reduced their invention to practice in November 1985. GE acknowledges that it did not reduce MAO to practice before Fonar. Therefore, in order to establish invalidity by anticipation, GE had to prove, by clear and convincing evidence, that it conceived the subject matter of the ’966 patent prior to November 1985, and, if so, that it exercised reasonable diligence until the time when it ultimately reduced the process to practice.

GE contends that, in April 1983, scientist Neil Hattes (“Hattes”) conceived GE’s MAO technique — the process upon which Fonar’s infringement charge is based. On direct examination, Hattes described a March 1983 entry in his notebook, which he had labeled “problem of oblique planes,” as his “initial statement of proposals, of ways that oblique plane imaging could be accomplished in a scan.” Trial Tr. at 1376. Next, he described a January 1984 entry in the same notebook, labeled “oblique plane musings,” as representing his “continuing thinking about more details about how we would build, what type of hardware would it take to perform the rotation.” Id. at 1382. Clearly such “proposals” and “musings” fall short of establishing conception.

Hattes also explained how he learned of a Texas Instruments microprocessor that could perform arithmetic at a high rate of speed, and how, in August 1984, he assigned to a GE engineer named Steven Deutsch (“Deutsch”) the task of “flushing out the design concept” for a board capable of performing a scan at an oblique angle. Id. at 1386-87. In November 1984, Deutsch finished an unpublished work, entitled “Micro-processing Implementation of MR Oblique Plane Imaging,” Deft Exh. 366. The Deutsch paper advised that certain software would be necessary “to accommodate multi-slice imaging where each slice may have a different orientation.” Id. It also found it necessary for such software “to provide an operator interface so that image rotation can be conveniently specified,” but indicated that the “details” of changes to GE’s existing MRI software were “beyond the scope of this report.” Id. Both Hattes and Deutsch testified that the Deutsch paper describes in detail a device for doing MAO. Trial Tr. at 1397-98, 1418-19.

Continuing on direct examination, Hattes testified that he assigned another GE engineer the task of designing hardware based on the Deutsch paper. What GE calls an Oblique Plane Imaging (“OPI”) board was completed January 1985. Id. at 1399. By February 1985, according to Hattes’s testimony, GE had the capacity to perform MAO by using the OPI board in conjunction with a dipswitch, a device used by the operator to select multiple imaging angles during the course of a single. Id. at 1403.

Evidence clearly establishes, however, that GE had not reduced MAO to practice in February 1985; in fact, it had yet to begin work on either the device driver or user interface software. An internal memorandum, entitled “Oblique Plane Imaging: An Initial Report,” which was prepared by Hattes for GE management in August 1985, listed “objectives” of GE’s “oblique plane imaging plan” followed by this text:

Future improvements may include:

• Multi-Angle, multi-slice (spine tracking) PLExh. 83. Indeed, Hattes’s use of the word “may” raises serious doubt as to whether he, or anyone else at GE, possessed in August 1985 a “definite and permanent idea of the complete and operative” MAO invention, as it would later be reduced to practice.

Nonetheless, GE insists that its scientists had conceived the MAO invention before August 1985. It proclaims that the Deutsch paper “set out a complete conception of GE’s multislice, multiangle imaging technique in November 1984.” Deft Mem. in Supp. of JMOL, at 16. “[A]ll that remained for GE’s commercial product was ... the routine software for interfacing with the hardware and to integrate it into the machine.” Id. at 17. Other than the oral testimony of GE’s own employees, GE did not present evidence supporting the notion that it had undertaken even preliminary work on the necessary software before November 1985, the date Fonar undeniably reduced to practice. Indeed, the outline of a presentation given by the lead software designer on February 17, 1986 indicates that GE still considered it an “objective” at that time to:

• Acheive (sic) Parity with Competitors

—Seimens has single angle about primary axis.

—Technicare has three angles prescription.

—Fonar allows each slice in a multi-slice to have its own angle.

PLExh. 65.

Moreover, the evidence falls short of establishing that GE’s OPI board was in fact a complete embodiment of the hardware necessary to perform MAO. Hattes acknowledged, on cross examination, that OPI modification would be necessary once the software was complete. Trial Tr. at 1472.

The Court determines that an inventor must have done at least some of the preliminary calculations necessary to design the proper software before he can claim to have “conceived” the subject matter of Claims 1, 2, 4, 5, or 12. For Claims 4 and 5, work must have been commenced on both the device driver and the user interface; for Claims 1 and 2, work must have been commenced on the device driver. The proof falls short of establishing that any software work was performed before November 1985.

The Court notes, however, that even if MAO was conceived by Hattes/Deutsch, GE failed to exercise reasonable diligence between the time of conception and the time of reduction to practice. GE suggests that MAO was conceived in November 1984, the time of the Deutsch paper, but GE admittedly did not reduce the invention to practice until September of 1986. GE should have been able to write the “routine” software— “all that remained for GE’s commercial product” — in less than twenty-two months.

The Court determines, as a matter of law, that GE was not the first to conceive the MAO technique, and that, even if it was, it failed to exercise reasonable diligence between conception and reduction to practice. There is substantial evidence in the record supporting the jury’s conclusion that GE failed to prove, by clear and convincing evidence, that the claims of the ’966 patent were anticipated by the work of Hattes et al. 8

4. Anticipation: The Work of Glover

GE asserts that Claims 4 and 5 of the ’966 patent are invalid because they were anticipated by the work of GE scientist Gary Glover, as memorialized by his report, published by GE in October 1984 (the “Glover reference”). See Deft Exh. 1222. A patent can be invalid as “anticipated” pursuant to 35 U.S.C. § 102(b) where the subject matter claimed by the patent was known or used by others in this country, or patented or described in a printed publication in this or a foreign country, before the invention by the patentee.

Glover gave the following testimony at trial. See Trial Tr. at 1127-1151. One of his responsibilities at GE was to perform magnet shimming. Magnet shimming is the process by which magnetic strength is maintained in a uniform fashion throughout an MRI scanner’s magnetic field. The tube-shaped magnet creates a magnetic field in the area inside the tube. At one time, shimming was performed by inserting wedge-shaped materials between the outside edge of the tube and its encasing. The shims would adjust the shape of the tube, and, thereby, adjust the balance of the magnetic field inside the tube. By trial and error the field could be balanced. Glover found this process to be entirely inefficient. He devised a process to collect MR data from scans taken at four different rotations in order to compute and measure the field strength for any given point in the field. The process produced images — usually of a bottle of water placed in the exact center of the field — the varying brightness of which would indicate strength of the field at a given point. Using the images, Glover could readily determine proper shim settings. Glover memorialized the process — which he called Chemical Shift Imaging (“CSI”) — in the Glover reference.

GE argues that the Glover reference anticipates Claim 4 of the ’966 patent. Anticipation, as indicated above, is a question of fact. As asserted here, it requires “the presence of a single prior art disclosure of each and every element of the claimed invention.” Electro Med. Sys. S.A. v. Cooper Life Sciences, 34 F.3d 1048, 1052 (Fed.Cir.1994). “The mere fact that a certain thing may result from a given set of circumstances is insufficient to prove anticipation.” Id. (emphasis in original). Rather, each claim limitation in the patent must necessarily be present in the prior art disclosure. Id.

As noted above, Claim 4 and dependent Claim 5 cover the use of manual cursor lines on a scout image to select desired imaging planes. The Glover reference neither contains the phrase scout image, nor discusses the concept of using a scout image as envisioned by Claim 4. See Deft Exh. 1222; Trial Tr. at 1176. On direct examination, Glover admitted that in performing CSI he “often took what we would call a scout image of what we called the bottle water to make sure it was in the right place to make sure before we did anything substantively.” Trial Tr. at 1151. He also stated that he used cursor lines. The evidence establishes that the scout image used by Glover and the scout image claimed by the ’966 patent are both to be used in a preliminary fashion. GE, however, did not establish that Glover used his scout scan as a means of selecting desired imaging planes. The evidence is insufficient to support the conclusion that Claims 4 and 5 of the ’966 patent are anticipated by the Glover reference, or Glover’s knowledge or use of CSI. There being no evidence upon which a reasonable jury could find that GE met its burden of proof, the Court determines, as a matter of law, that the ’966 patent is not invalid as anticipated by the work of Glover.

5. Obviousness

Even where the patented invention is not “anticipated” pursuant to § 102, it may be deemed invalid pursuant to 35 U.S.C. § 103 if the differences between the subject matter of the claimed invention and the prior art are such that the subject matter as a whole would have been “obvious” to a person of ordinary skill in the art at the time the invention was made.

Obviousness is a question of law, based on underlying factual inquiry. Panduit Corp. v. Dennison Mfg. Co., 810 F.2d 1561, 1566-68 (Fed.Cir.), cert. denied, 481 U.S. 1052, 107 S.Ct. 2187, 95 L.Ed.2d 843 (1987). The ultimate legal conclusion is whether the subject matter of a claimed invention would have been obvious to one skilled in the prior art. Id. at 1568. The factual inquiry involves consideration of the following factors set forth in Graham v. John Deere Co., 383 U.S. 1, 17-18, 86 S.Ct. 684, 693-694, 15 L.Ed.2d 545 (1966): (1) the scope and content of the prior art; (2) the differences between the prior art and the claims; (3) the level of ordinary skill in the art at the time of the invention; and (4) objective evidence of nonobviousness, such as long-felt but unsatisfied need, commercial success, and copying. The degree to which the obviousness question is one of fact is the “degree required to erect a foundation of facts capable of supporting the conclusion.” Panduit, 810 F.2d at 1568. The conclusion is the legal question the court must answer: “whether the patent challenger carried its burden of establishing invalidity.” See id. at 1569.

The jury returned negative answers as to Claims 1, 2, 4, 5, and 12 in response to an interrogatory asking whether the subject matter ’966 patent would have been “obvious, in light of the prior art, to one skilled in the art at the time the invention was made.” On GE’s motion, the Court must determine whether there was substantial evidence upon which a reasonable jury could have found facts consistent with its conclusion of nonobviousness.

Fonar contends that only those references disclosed to the United States Patent and Trademark Office qualify as prior art for the purpose of the obviousness determination. GE insists, however, that the knowledge, work and/or devices of Hattes and Glover should be included, as well. See Trial Tr. at 1807. As explained below, the subject matter of the ’966 patent would not have been “obvious,” even if the prior art included the work of Hattes and Glover.

With the exception of the Deutsch paper, GE’s evidence supporting the proposition that the Hattes conception was prior art is a collection of oral testimony from various GE personnel. Because “oral testimony to establish the existence of allegedly anticipatory devices has long been viewed with skepticism,” Sjolund v. Musland, 847 F.2d 1573, 1578 (Fed.Cir.1988), a reasonable jury could certainly have found that — with the exception of the Deutsch paper — the subject matter allegedly conceived by Hattes was not prior art. Assuming, arguendo, that the jury did consider the Deutsch paper prior art, differences between the prior art and the claims of the ’966 patent would have been the software necessary to perform MAO.

As stated above, two types of software are necessary to perform MAO — device drivers and graphic prescription. As to graphic prescription, the evidence is conflicting. In support of its contention that graphic prescription was an obvious leap from the hardware, GE points to the testimony of Smith. He was questioned, as follows, about an MRI scanner that he used prior to the invention of MAO, the various capacities of which had been disclosed in the ’966 patent as prior art:

Q. As far as using that cursor then to do multiple slices at different angles, say two slices at different angles, was that an obvious change to what you had on the machine that was there in the summer of ’85, just have the cursor move from one angle to another and it would record those two angles?

A. I think that was a pretty clear extension of what we were doing. That is not something we labored long over.

THE COURT: Was it obvious to someone skilled in the art to go from one to various lines?

A. I think in terms of the means of how to set up the positioning of the different slices it’s a pretty obvious extension of how we did it.

Trial Tr. at 1270-71. Fonar, however, identifies the following testimony of Hattes in support of the contrary view:

Q. [Graphic prescription is] a user interface. Would that be the correct way to describe it?

A. It is a user interface.

Q. And as used here on the image on the screen, it’s the user interface that tells the computer — its the software that tells the computer which planes are to be imaged and at what angles, correct?

A. Well, I think it’s the start of a very complex process from the drawing to some resultant information that would lead to the scan that would then be requested.

Q. Why did you call that a complex process?

A. Well, I think that the interpretation of where the lines are, what that might mean in terms of what the system has to follow, those directions, in my mind, was not an easy follow from one to the other.

Id. at 1469-70. Although Hattes’s expertise is not software design, a reasonable jury could have found him to represent one skilled in the art at the time of the invention. The same holds true for Smith, although his testimony is particularly compelling because he is an inventor on the ’966 patent.

Despite Smith’s testimony, the Court determines that, even if the jury considered the Deutsch reference prior art, there is substantial evidence to support factual findings that could serve as a basis for a conclusion of nonobviousness. This determination is based, not only on the testimony of Hattes, but also on GE’s failure to produce evidence supporting the obviousness of device driver software. In addition, the substance of GE’s best mode and enablement arguments undermine the force of its obviousness assertions. For example, GE’s “patent” expert, Joseph Colaianni (“Colaianni”), testified that “the disclosure in the [’966] patent with regard to the software” was insufficient because it did not contain “a complete description of the invention so as to enable a person skilled in the art to make and use it.” Id. at 1784. On the one hand, GE contends that one skilled in the art could not have written the necessary software because the ’966 patent did not sufficiently disclose how to write it; yet, on the other hand, GE insists that to one skilled in the prior art designing the software would be “obvious.” The former contention diminishes the viability of the latter. If one cannot design the necessary software based on the prior art and the ’966 patent disclosure, it cannot then be an “obvious” task for one to design the same software based on the prior art without the ’966 patent disclosure.

GE also makes an obviousness argument in light of Glover’s work. The Glover reference, however, adds little to GE’s position. Based on the jury’s implied factual findings — which are supported by substantial evidence even if both the Deutsch and the Glover references were considered prior art — the Court determines, as a matter of law, that GE failed to prove, by clear and convincing evidence, that the subject matter of Claims 1, 2, 4, 5, and 12 would have been obvious at the time of the invention to one skilled in prior art.

D. Inducement of Infringement

The jury found that GE infringed Claims 1, 2, 4, 5, and 12 of the ’966 patent, and the Court determines, without need for discussion, that the jury’s finding is supported by substantial evidence. Separate from direct infringement under 35 U.S.C. § 271(a), a party’s acts in connection with sale of equipment may constitute active inducement of infringement under § 271(b). See Joy Technologies., Inc. v. Flakt, Inc., 6 F.3d 770, 774 (Fed.Cir.1993). “Proof of actual intent to cause the acts which constitute infringement is a necessary prerequisite to finding active inducement.” Hewlett-Packard Co. v. Bausch & Lomb Inc., 909 F.2d 1464, 1469 (Fed.Cir.1990). In addition, liability for active inducement is dependent on and “cannot occur without an act of direct infringement.” Joy Technologies, 6 F.3d at 774.

One cannot infringe without notice of infringement. Fonar concedes that GE is not liable for direct infringement on MRI sales occurring before September 1992, the date when Fonar gave GE notice of infringement. Fonar seeks lost profit damages, however, for sales made prior to September 1992 on the theory that GE’s service and repair after September 1992, of scanners sold prior to September 1992, constitutes active inducement. Even though the Court did not charge the jury on active inducement, it asked the jury to determine damages for active inducement. The jury awarded $13,-625,000.

Fonar asserts that its active inducement theory finds support in Aro Mfg. Co v. Convertible Top Replacement Co., 377 U.S. 476, 484, 84 S.Ct. 1526, 1531, 12 L.Ed.2d 457 (1963), wherein the Supreme Court instructed that “[w]here use infringes, repair does also, for it perpetuates the infringing use.” As the quoted language implies, however, and as the law of active inducement clearly states, there must be an underlying act of direct infringement before there can be an act of inducement of infringement. Indeed, the Supreme Court’s next sentence leaves little room for doubt. It explained: “ ‘a patented article may be repaired without making the repairer an infringer[ ].... where the device in patented form has come lawfully into the hands of the person for whom it is repaired.’ ” Id. (quoting Union Special Mach. Co. v. Maimin, 161 F. 748, 750 (C.C.E.D.Pa.), aff'd, 165 F. 440 (C.A. 3d Cir. 1908)). GE was not liable for direct infringement on sales occurring prior to September 1992. With no underlying direct infringement as to those scanners, there can be no active inducement.

In retrospect, the jury should not have received an interrogatory on active inducement. The interrogatory was given at Fo-nar’s request and over GE’s objection. Having reviewed the law, however, the Court is convinced that Fonar’s theory has no basis. Nevertheless, even if it did, no reasonable jury could have concluded that Fonar proved GE’s actual intent to induce infringement. As such, the Court must disregard the jury’s answer to question 5 on Special Verdict Sheet #2.

E. Damages

A patentee can recover damages for infringement, pursuant to 35 U.S.C. § 284, under two theories: lost profits and reasonable royalty. See Rite-Hite Corp. v. Kelley Co., 56 F.3d 1538, 1544 (Fed.Cir.1995) (en banc). Lost profits are available where the record permits an accurate determination of the patentee’s loss of sales and profits because of the infringement. Reasonable royalties are always available. An award may be split between lost profits, as actual damages to the extent they are proved, and reasonable royalties for the remainder. State Indus., Inc. v. Mor-Flo Indus., Inc., 883 F.2d 1573, 1577 (Fed.Cir.1989), cert. denied, 493 U.S. 1022, 110 S.Ct. 725, 107 L.Ed.2d 744 (1990). Determining the nature and amount of damages is a question of fact on which the patentee has the burden of proof BIC Leisure Prods. v. Windsurfing Int'l, 1 F.3d 1214, 1217 (Fed.Cir.1993).

1. Lost Profits

GE moves the Court to disregard the jury’s finding of $27,825,000 in lost profits, based on 75 “lost” sales, as not supported by substantial evidence. Fonar contends that there is evidence in the record upon which a reasonable jury could base such a conclusion.

In order to recover lost profits, “the patentee must show a reasonable probability that, ‘but for’ the infringement, it would have made the sales that were made by the infringer.” Rite-Hite, 56 F.3d at 1545 (quoting Del Mar Avionics, Inc. v. Quinton Instrument Co., 836 F.2d 1320, 1326 (Fed.Cir.1987)). The patentee must prove its claim to lost profits by a preponderance of the evidence. Yarway Corp. v. Eur-Control USA, Inc., 775 F.2d 268, 275 (Fed.Cir.1985). This burden includes the presentation of proper evidence for the computation of the lost profits. Standard Havens Prods. v. Gencor Industs., 953 F.2d 1360, 1372 (Fed.Cir.1991) (citing King Instrument Corp. v. Otari Corp., 767 F.2d 853, 863 (Fed.Cir.1985), cert. denied, 475 U.S. 1016, 106 S.Ct. 1197, 89 L.Ed.2d 312 (1986)), cert. denied, — U.S. -, 113 S.Ct. 60, 121 L.Ed.2d 28 (1992).

Because the determination of damages “is not an exact science, ... ‘the amount need not be proven with unerring precision.’ ” Del Mar Avionics, Inc. v. Quinton Instrument Co., 836 F.2d 1320, 1327 (Fed.Cir.1987) (quoting Bio-Rad Labs., Inc. v. Nicolet Instrument Corp., 739 F.2d 604, 616 (Fed.Cir.), cert. denied, 469 U.S. 1038, 105 S.Ct. 516, 83 L.Ed.2d 405 (1984)). In addition, “[w]hen the amount of the damages cannot be ascertained with precision, all doubts are resolved against the infringer.” Id.

A useful test for determining whether a patentee met the “but for” requirement was enunciated in Panduit Corp. v. Stahlin Bros., 575 F.2d 1152 (6th Cir.1978). This test requires the patentee to establish the following four factors: (1) demand for the patented product; (2) an absence of acceptable noninfringing substitutes; (3) the patentee’s ability to exploit the demand; and (4) the amount of profits the patentee would have made. Id. at 1156. The patentee’s burden, however, is to prove no more than a reasonable possibility that the sales would have been made; the patentee need not negate “every possibility” that the purchaser might have bought another noninfringing product. Rite-Hite, 56 F.3d at 1545. The Court determined that it was unnecessary to charge the jury on the Panduit test.

According to the cross-examination testimony of Jay D. Miller (“Miller”), whom GE employs as the Americas Marketing Manager for MRI systems, Fonar and GE directly competed in the market for MRI scanners with magnetic field strengths of .3 to .5 Tesla. Trial Tr. at 2189. In that particular market, Miller testified that GE sold 21 Vec-tra scanners and “about” 50 Spectra scanners after September 1992. Id. Fonar acknowledges that there were at least four other manufacturers competing in the market. Id. at 2116. The testimony of economics expert Laurits Christensen (“Christensen”) indicated, however, that Fonar considered all manufacturers in the market to be infringers of the ’966 patent and, therefore, considered every sale in the market to be a lost sale. Id.

Fonar presented evidence that it would have been able to meet demand for scanners equipped with MAO. Damadian testified that on the average Fonar could manufacture three scanners a month, and that, in 1988, at its peak rate of production, it was capable of manufacturing eight per month. Id. at 2049. Taking into account the corporation’s growth rate, he added that Fonar would have been able in 1992 to manufacture 500 MRI scanners a year. Id. at 2050.

Viewing the record as a whole and drawing all inferences in favor of Fonar, the Court determines that there is substantial evidence supporting the jury’s conclusion that Fonar had reasonable possibility of selling 75 scanners but for GE’s infringement. GE’s motion is denied in this respect. Because the record also supports the jury’s determination as to the amount of lost profit damage, the Court denies GE’s request to reduce the amount of the award.

2. Reasonable Royalties

Where, as in the instant case, there is no established royalty between the parties, a reasonable royalty should be based upon “the supposed result of hypothetical negotiations between the plaintiff and defendant.” Rite-Hite, 56 F.3d at 1554. “The hypothetical negotiation requires the [jury] to envision the terms of a licensing agreement reached ... at the time of infringement.” Id. A defendant challenging a jury’s determination, must show that, in view of all the evidence, the award is “ ‘so outrageously high ... as to be unsupportable as an estimation of a reasonable royalty.’ ” Id. (quoting Lindemann Maschinenfabrik GmbH v. American Hoist & Derrick Co., 895 F.2d 1403, 1406 (Fed.Cir.1990)).

After awarding Fonar lost profits on 75 of the 600 infringing scanner sales, the jury added $34,125,000 as a reasonable royalty on the remaining 525 infringing units. GE contends that the award is outrageously high. Fonar supports the award as reasonable in light of the evidence, particularly the testimony of Christensen.

Christensen testified that a reasonable royalty could be a percentage of GE’s total revenues from the sales of infringing scanners. Trial Tr. at 2079. Christensen stated that a known “rule of thumb” for determining an appropriate royalty rate was one-quarter to one-third of anticipated profits on the sale of infringing scanners. Id. Christensen then stated that, according to GE’s records, GE earned a 25 percent operating profit on infringing sales. Id. at 2079-80. Given that, he determined that 7.25 percent — “the midpoint” — would be a reasonable royalty rate. Id. at 2080. Christensen then calculated a $54 million reasonable royalty for the 525 infringing sales.

Raymond S. Sims (“Sims”), an economist, and George R. Clark (“Clark”), a patent licensing specialist, testified on GE’s behalf. Sims indicated that a reasonable royalty would be between $1,000 and $3,500 per unit. Id. at 2214-19. Clark indicated that $1,000 per unit would be appropriate. Id. at 2285. The decisive factor underlying both experts’ conclusions was Fonar’s acknowledgment that it had sold MAO as a separate and distinct MRI feature for $1,000 to $1,500, and as part of an upgrade package worth $15,000. GE, therefore, suggests that a reasonable royalty on the remaining 525 infringing sales would be in the $525,000 to $1,837,500 range.

The gap between the experts’ evaluations raises the issue of whether Fonar is entitled to'a royalty based on the entire value of an MRI scanner, or merely the value of MAO as but one feature of the scanner. Under the “entire market value” rule, damages are recoverable on the value a paten-tee’s entire apparatus containing several un-patented features where “the patent-related feature is the ‘basis for customer demand.’ ” Rite-Hite, 56 F.3d at 1549-50 (quoting State Indus., 883 F.2d at 1580)). Where the patented and unpatented components of the device are physically part of the same machine, the patentee can typically recover compensation based on the value of the entire apparatus. See id. (citing Western Elec. Co. v. Stewart-Warner Corp., 631 F.2d 333 (4th Cir.1980), cert. denied, 450 U.S. 971, 101 S.Ct. 1492, 67 L.Ed.2d 622 (1981)). Nonetheless, the Court instructed the jury that it could “consider whether it is appropriate to base compensation on the entire apparatus normally anticipated to be sold, even though only one feature is patented.” Trial Tr. at 2448. The size of the jury’s reasonable royalty award indicates that it chose not to accept GE’s theory of valuation. Based on Rite-Hite, that choice was a reasonable one.

The jury was also instructed that, in determining a reasonable royalty, it should consider “how much profit or cost savings the infringer expected to realize from his practice of the [patented] invention” at the time of infringement. Id. at 2449-50. Miller, GE’s MRI marketing specialist, testified on direct examination as to the decline in demand for MAO since 1992. He stated that “[t]hese days, multi-slice multi-angle is not as an important feature as it was at one time.” Id. at 2168. He explained that “before 1992,.... [y]ou wanted to image as many patients per day as possible and there was actually a waiting list of patients to come in and scan if you had an extra hour or two per day.” Id. at 2169. Miller’s testimony indicates, quite clearly, that demand for MAO was high at the time when “hypothetical negotiations” would have taken place. Based on the testimony of its own marketing expert, the jury could have reasonably concluded that GE would have paid a royalty reflecting the high demand for MAO at the time of infringement.

The Court determines that substantial evidence supports the reasonableness of the jury’s royalty award. GE’s motion is denied in this regard. For the same reason, the Court denies GE’s request for remittitur.

3. Injunction

Based on the foregoing, GE and South Shore and their officers, agents, servants, and employees, and those persons in active concert or participation with them, are restrained and enjoined, from and after the date of this Memorandum & Order and until the expiration of the ’966 patent or the date upon which any court enters an order (1) vacating the jury’s finding of infringement of the claims of the ’966 patent or (2) finding the claims of the ’966 patent invalid, from infringing, directly, indirectly, contributorily, or by inducement, the ’966 patent by making, using, or selling, or causing to be made, used, or sold any MRI scanners.

Further, GE and South Shore and their officers, agents, servants, and employees, and those persons in active concert or participation with them, are restrained and enjoined from and after the date of this Memorandum & Order and until the expiration of the ’966 patent or the date upon which any court enters an order (1) vacating the jury’s finding of infringement of the ’966 patent or (2) finding the claims of the ’966 patent invalid from infringing, directly, indirectly, con-tributorily, or by inducement, the ’966 patent by making, using, or selling, or causing to be made, used, or sold any upgrade of software and/or hardware to any MRI scanner that would give such MRI scanner the capability of obtaining in the course of a single scan NMR image data from a plurality of differently oriented selected planes in an object using magnetic resonance techniques in accordance with the claims of the ’966 patent.

Nothing in this order shall preclude GE or South Shore and their officers, agents, servants, and employees, and those persons in active concert or participation with them from using, servicing, or upgrading in any manner whatsoever any MRI scanner delivered by GE before May 26, 1995, and such using, servicing, or upgrading shall not create additional liability under the ’966 patent.

4. Interest

Under 35 U.S.C. § 284, a judgment may be accompanied by “interest and costs as fixed by the court.” The Supreme Court has stated that “prejudgment interest should be awarded under § 284 absent some justification for withholding such an award.” General Motors Corp. v. Devex Corp., 461 U.S. 648, 657, 103 S.Ct. 2058, 2063, 76 L.Ed.2d 211 (1983). Prejudgment interest serves to compensate the patentee because, in addition to lost profits and royalties payments, he also lost the use of that money between the time of infringement and the date of judgment. Id. A court has substantial discretion to determine the interest rate in patent eases. Gyromat Corp. v. Champion Spark Plug Corp., 735 F.2d 549, 556-57 (Fed.Cir.1984).

The parties each presented intricate arguments, supported by detailed expert affidavits, and somehow reached starkly different results. The Court finds that the record lacks support for many of the contentions advanced.

The Court determines that fair and adequate compensation includes an award of prejudgment interest. The rate of interest should be based on the average 52-week United States Treasury bill rate for the period of September 1992 to the date of this Memorandum & Order. Prejudgment interest shall be computed daily and compounded annually.

III. CONCLUSION

For the above reasons, GE’s motion is granted in part and denied in part. The motion is granted insofar as it seeks entry of JMOL in favor of GE on grounds that it did not directly infringe the ’832 patent and did not induce infringement of the ’966 patent. The balance of the motion for JMOL, like GE’s motion for a new trial, is denied in all respects. Insofar as this Memorandum & Order grants injunctive relief, however, it is stayed pending appeal, subject to a bond.

Counts II, IV, V, VI, VII, and VIII of the complaint having been withdrawn, and the Hitachi defendants having been released from this action, the Clerk of the Court is directed to enter judgment consistent with this Memorandum & Order and as follows:

(1) as to Court I of the complaint, defendants GE and South Shore have not directly infringed or induced infringement of Claims 1 and 2 of the United States Patent No. 3,789,832;

(2) defendants GE and South Shore have failed to establish any of their defenses as to United States Patent No. 3,789,832;

(3) Count I of the Complaint is dismissed;

(4) as to Count III of the complaint defendants GE and South Shore, by jury verdict, have directly infringed Claims 1, 2, 4, 5, and 12 of United States Patent No. 4,871,966;

(5) as to Count III of the complaint, defendant GE has not induced infringement of Claims 1, 2, 4, 5, and 12 of United States Patent No. 4,871,966;

(6) defendants GE and South Shore have failed to establish any of their defenses as to United States Patent No. 4,871,966; and

(7) as,to Count III of the complaint, by jury verdict, plaintiffs Fonar and Raymond Y. Damadian are awarded and defendants GE and South Shore must pay a sum of $61,950,000, plus prejudgment interest as set forth above.

SO ORDERED.

APPENDIX

The claims of the ’966 patent are as follows:

1. A method for obtaining in the course of a single scan NMR image date for a plurality of differently oriented selected planes in an object using nuclear magnetic resonance techniques, said method comprising the steps of:

(a) positioning an object in a static homogeneous magnetic field;

(b) determining first and second selected planes in said object for which NMR image data is to be obtained, said first selected plane being located at a first portion of said object and having a first orientation with respect to a predetermined direction and said second selected plane being located at a second portion of said object and having a second orientation with respect to said predetermined direction, said first and second orientations being different from one another;

(c) subjecting said object to a plurality of repetitions of a first repetition sequence composed of NMR excitation and magnetic gradient field pulses, each of said repetitions of said first repetition sequence including the steps of applying an excitation pulse and reading out of an NMR signal produced by said excitation pulse, said excitation pulse for said first repetition sequence being applied at a first predetermined frequency in the presence of a first predetermined slice selector magnetic field gradient having a gradient direction extending perpendicular to said first selected plane, said first predetermined frequency being chosen to that said application of said excitation pulse at said first predetermined frequency only excites selected nuclei in said first selected plane, and said plurality of repetitions of said first repetition sequence being carried out in a manner to encode spatial information into a first collection of said NMR signals, said first collection of NMR signals being representative of NMR image data for said first selected plane; and

(d) subjecting said object to a plurality of repetitions of a second repetition sequence composed of NMR excitation and magnetic field gradient pulses, each of said repetitions of said second repetition sequence including the steps of applying an excitation pulse and reading out of an NMR signal produced, by said excitation pulse, said excitation pulse for said second repetition sequence being applied at a second predetermined frequency in the presence of a second predetermined slice selector magnetic field gradient having a gradient direction extending perpendicular to said second selected plane, said second predetermined frequency being chosen so that said application of said excitation -pulse at said second predetermined frequency only excites selected nuclei in said second selected plane, and said second predetermined slice selector magnetic field gradient and said second predetermined frequency being different from said first predetermined slice selector magnetic field gradient and said first predetermined frequency, respectively, and said plurality of repetitions of said second repetition sequence being carried out in a manner to encode spatial information into a second collection of NMR signals, said second collection of NMR signals being representative of NMR image data for said second selected plane; said plurality of repetitions of said first and second repetition sequences each being carried out during the course of a single scan of said object and each being continued substantially throughout said single scan, the repetition time interval for repeating each of said first and second repetition sequences being substantially the same and said steps of applying an excitation pulse and reading out of an NMR signal for each repetition of said second repetition sequence being performed at a different time during said repetition time interval than each of said steps of applying an excitation pulse and reading out of an NMR signal for said first repetition sequence.

2. The method of claim 1, wherein step (c) comprises generating said magnetic gradient field pulses of said first repetition sequence via a first and a second wave-form, corresponding to said first orientation of said first selected plane, which produce said first predetermined slice selector magnetic field gradient, and a first predetermined read out magnetic field gradient having a direction orthogonal to that of said first predetermined slice selector magnetic field gradient; and

wherein step (d) comprises generating said magnetic field

gradient pulses of said second repetition sequence via a third and a fourth waveform, corresponding to said second orientation of said second plane, which produce said second predetermined slice selector magnetic field gradient, and a second predetermined read out magnetic field gradient having a direction orthogonal to that of said second predetermined slice selector magnetic field gradient.

4. A method for conducting an examination of an object along two different selected planes using nuclear magnetic resonance techniques, said method comprising the steps of:

(a) positioning an object in an NMR imaging apparatus which includes means for generating a magnetic field, means for exciting selected nuclei to generate NMR signals and for reading of such NMR signals to provide a collection of NMR signals from selected regions of an object placed in said NMR imaging apparatus, means for applying gradient magnetic fields, means for obtaining NMR imaging date from said collection of NMR signals and means for producing an image from said NMR imaging date;

(b) operating said NMR imaging apparatus to obtain an NMR scout image for a portion of said object of said examination:

(c) while said object remains positioned in said NMR imaging apparatus, using said scout image to select a first plane and a second plane of said object for which NMR image data is to be obtained, said first plane and said second plane each being transverse to said scout plane, and said first plane having a first orientation relative to said scout plane and said second plane having a second orientation relative to said scout plane, said first orientation being different from said second orientation;

(d) conducting a plurality of NMR sampling operations to obtain NMR imaging date from said first selected plane of said object, said step of conducting said sampling operations for said first plane being commenced at a first time during the course of a single scan and being continued substantially through out said single scan so as to obtain NMR imaging date for said first-selected plane of said object;

(e) conducting a plurality of NMR sampling operations to obtain NMR imaging data from said second selected plane of said object which is different from said first selected plane, said step of conducting said sampling operations for said second selected plane being commenced at a second time during the course of said single scan which is later than said first time, but prior to completion of said step of conducting said sampling operations for said first selected plane, and said step of conducting said NMR sampling operations for said second selected plane being continued substantially throughout said single scan to obtain NMR imaging data for said second selected plane; each of said plurality of NMR sampling operations

including an NMR excitation operation and an NMR reading operation, said NMR excitation operations for each of said selected planes being carried out in a manner so as to excite selected nuclei in said each of said selected planes, and said NMR imaging data, each of said NMR excitation and NMR reading operations being performed at a different time during the course of said single scan than each other of said excitation and reading operations.

5. The method of claim 4, wherein step (d) comprises applying a first and a second waveform, corresponding to said first orientation of said first plane, to said means for applying gradient magnetic fields, to produce a first predetermined slice selector magnetic field gradient having a direction orthogonal to said first plane, and wherein;

step (e) comprises applying a third and a fourth waveform,

corresponding to said second orientation of said second plane, to said means for applying gradient magnetic fields, to produce a second predetermined slice selector magnetic field gradient having a direction orthogonal to said second plane.

12. Apparatus for obtaining NMR image data from a plurality of selected planes in an object comprising:

(a) means for applying magnetic fields to the object;

(b) means for applying radio frequency excitation pulses to the object;

(c) means for actuating and controlling said magnetic field applying means and said radio frequency applying means to:

(1) apply a first sequence including a first slice selector magnetic field gradient in a first direction concomitantly with a first RF excitation pulse at a first frequency to thereby excite nucleii only in a first plane perpendicular to said first direction, whereby a first NMR signal will be emitted only by nucleii in said first plane, said first sequence further including at least one encoding magnetic field gradient operative to encode spatial information into said NMR signal;

(2) apply a second sequence including a second slice selector magnetic field gradient in a second direction different from said first direction concomitantly with a second RF excitation pulse at a second frequency different from said first frequency to thereby excite nucleii only in a second plane perpendicular to said second direction whereby a second NMR signal will be emitted only by nucleii in said second plane, said second sequence further including at least one encoding magnetic field gradient operative to encode spatial information into said second NMR signal; and

(3) a repeat said first and second sequences a plurality of times during the course of a single scan so as to excite nucleii in said first and second planes alternatively and produce said first and second NMR signals alternately while varying said at least one encoding gradient in each said sequence on each repetition; and

(d) means for receiving said first and second NMR signals and recovering NMR image data therefrom.

1 . The complaint also named as defendants Hitachi LTD, Hitachi Medical Corporation, Hitachi Medical Systems America, Inc., Summit World Trade Corporation, Hitachi Medical Corporation of America, and Brookhaven Magnetic Resonance Imaging, Inc. (the "Hitachi defendants"). Upon settlement of the claims asserted against them, the Hitachi defendants were released from the action prior to trial, pursuant to Stipulation and Order of Dismissal with Prejudice dated May 18, -1995.

2 . The following discussion on NMR technology is based on the testimony of Robert G. Peyster, M.D. ("Peyster”), an independent expert appointed by the Court with the parties’ consent. Also with the parties' consent, Peyster's testimony was evidence in the instant case.

3 . Although it is permissible under Markman to construe the claims on a post-verdict motion for JMOL, the Court acknowledges that the course it has taken in this case cannot be preferred. Indeed, providing jurors with the proper interpretation before they deliberate makes more sense. In a perfect world, the Court would have included a definition of the claims in its charge.

The subject matter underlying this case, however, was exceedingly complex. The Court endeavored for months to enlist the assistance of an independent expert. The parties would not consent because, by their estimation, nearly every expert in the field was linked in some capacity to one of the parties. On the eve of trial the parties did consent to having Peyster read a pre-ap-proved, joint statement to the jury regarding MRI technology and United States Patent and Trademark Office procedure.

Also on the eve of trial, the Markman decision came down from the United States Court of Appeals for the Federal Circuit. If the Markman decision had been rendered earlier, the Court would have insisted on having expert assistance from the outset. Only then would it have been possible for the Court to acquire a complete understanding of the technologies involved in time to charge the jury with a construction of the claims.

As much as it might have seemed the more logical course, it would neither have been wise nor proper for the Court to instruct the jury as to the meaning of the claims on short notice. A task of such grave importance must be undertaken after long and careful review. The Court determined that it was best to reserve claim construction until that time.

4 . Indeed, the Court could have granted GE's motion for JMOL at the close of Fonar's case on the ground that the properly construed claims did not read on imaging as performed by GE's scanners, but the Court determined that the safer course was to allow all evidence to be presented and all arguments to be made before rendering an interpretation of the claims.

5 . Even though, having reviewed the evidence, the Court does not find strong support for the proposition that GE proved, by clear and convincing evidence, that the inventors on the '966 patent did, in fact, contemplate a best mode for carrying out their invention at the time of filing, the Court treats the following language from Fonar's memorandum of law submitted in oppo sition to GE’s motion for JMOL as an acknowl-edgement that a best mode was contemplated. The language is: "[t]he evidence makes clear that at the time the application for the MAO patent was filed, Fonar went to great lengths to disclose the best mode it knew.” Pl.Mem., at 23.

6 . The full text to which Wolf refers reads: "Further, in response to the coordinate and angle data associated with each slice output by the system 70, the computer 71 calculates for each slice an RF excitation frequency for the plane orthogonal to the direction of the slice selector gradient rotated by the angle of the slice, which passes through the selected portion of the object, and a frequency required to demodulate the corresponding NMR signal during the read out period.”

7 . GE's claim that the best mode requirement is violated merely because Fonar maintains as trade secrets certain aspects of the hardware and software has no merit. See, e.g., In re Hayes, 982 F.2d at 1538 (holding that best mode was adequately disclosed even where a patentee held as a trade secret the fact that a particular implementation of his invention was in firmware).

8 . The jury was instructed that it could find a patent invalid where the “inventor did not himself invent the subject matter sought to be patented.” Trial Tr. at 1969. In addition, the jury was instructed: [i]n determining priority of invention there shall be considered not only the respective dates of conception and reduction to practice of the invention, but also the reasonable diligence of one who was the first to conceive and the last to reduce to practice.” Id. In light of this instruction, the jury's affirmative response to interrogatory II.C., which asked whether "Drs. Smith, Hertz, Wolf, and Olsen, by themselves, invent[ed] the subject matter claimed,” was a finding that the '966 patent was not anticipated by prior invention.