Appeal from a judgment of the District Court of Minnesota holding U.S. Patents 3,391,697 (Greatbatch), 3,833,005 (Wingrove) and 3,901,247 (Walmsley) invalid, and finding no infringement. We modify and affirm.

Medtronic, Inc. and Med-Rel, Inc. (Medtronic) own the asserted patents directed to implantable cardiac pacemakers. Greatbatch describes and claims a pacemaker which is "runaway inhibited", i.e., one that does not stimulate the heart above a predetermined rate. Wingrove describes and claims a digital programmable pacemaker which can be reprogrammed by radio frequency signals after implantation. Walmsley describes and claims a pacemaker having an easily interpreted indication of battery condition.

Medtronic sued Cardiac Pacemakers, Inc. (CPI) for infringement of Greatbatch claims 1, 8, 10, 11, and 12, Wingrove claims 1, 4, 5, 7, 8, 9, 10, and 23, and Walmsley claim 13. CPI denied infringement and counterclaimed, asserting that the patents were invalid because the inventions of Greatbatch and Wingrove lacked novelty and because all the claimed inventions would have been obvious.

On January 19, 1983, Judge Devitt issued an 11 page Memorandum and Order, published at 555 F. Supp. 1214, finding the claimed inventions of Greatbatch and Wingrove novel, and holding that the "runaway inhibit system" of Greatbatch, the "remotely programmable, digitally timed and controlled pacer" of Wingrove, and "the Walmsley patent" would have been obvious under 35 U.S.C. Sec. 103.1 Though the Memorandum says the court did not "reach the question of infringement", it includes a finding that the claims at issue are not infringed, apparently on the theory that invalid claims cannot be the basis of liability for infringement.

The dispositive issue is whether the district court correctly held that the inventions before it, as set forth in the asserted claims of the Greatbatch, Wingrove, and Walmsley patents, would have been obvious under Sec. 103.2

Medtronic's appeal rests on assertions that the findings underlying the appealed judgment are clearly erroneous and on the contention that words and phrases appearing in the Memorandum establish the presence of fundamental errors of law in ignoring and misapplying the statutory presumption of validity provided by 35 U.S.C. Sec. 282,3 and in applying an improper test for determining the obviousness/nonobviousness issue. Though the Memorandum, as discussed below, contains language and phrases unsupported by the statute, Title 35 U.S.C., that alone does not require reversal where, as here, the decisional approach has not been shown to have been influenced or controlled thereby, and the result of the trial has not been shown, on the entire record, to have been reached on the basis of either clearly erroneous findings or on an actual misapplication of the law.

(1) The Memorandum and Order

Medtronic vigorously argues error in the Memorandum statement that "[c]ourts should scrutinize combination patent claims with a care proportioned to the difficulty and improbability of finding invention in an assembly of old elements...." We cannot construe that statement, taken from the opinion in Sakraida v. Ag Pro, Inc., 425 U.S. 273, 281, 96 S. Ct. 1532, 1537, 47 L. Ed. 2d 784 (1976) and Great Atlantic & Pacific Tea Co. v. Supermarket Equipment Corp., 340 U.S. 147, 152, 71 S. Ct. 127, 130, 95 L. Ed. 162 (1950) as a rule of law applicable broadly to patent cases because virtually every claimed invention is a combination of old elements, and, as Judge Devitt noted, virtually every patent can be described as a "combination patent". Stratoflex, Inc. v. Aeroquip Corp., 713 F.2d 1530, 218 U.S.P.Q. (BNA) 871, 880 (Fed.Cir.1983). Further, the Court has itself recognized the patentability of combinations of old elements. United States v. Adams, 383 U.S. 39, 51-52, 86 S. Ct. 708, 714-715, 15 L. Ed. 2d 572, 148 U.S.P.Q. (BNA) 479, 483 (1966) ("Despite the fact that each of the elements of the Adams battery was well known in the prior art, to combine them as did Adams required that a person reasonably skilled in the prior art must ignore" long-accepted factors.).

There is neither a statutory distinction between "combination patents" and some other, never defined type of patent, nor a reason to treat the conditions for patentability differently with respect to "combination patents". It but obfuscates the law to posit a non-statutory, judge-created classification labeled "combination patents". Richdel, Inc. v. Sunspool Corp., 714 F.2d 1573, 219 U.S.P.Q. (BNA) 8, 12 (Fed.Cir.1983). That the misstatement did not in this case lead to an erroneous application of the law, however, is indicated in part by the Memorandum statement that the "elements presumptively were combined in an inventive manner so as to render the combination patentable". Though the record would have been more clear without the reference to "combination patents", and "inventive manner", it cannot be said that the appealed judgment was in this case so influenced thereby as to require reversal.

Medtronic's contention that the statutory presumption of validity was misapplied is based on the Memorandum statement that the "presumption of validity is weakened, if not completely destroyed, by proof of pertinent prior non-considered art". The district court was there relying on language appearing in circuit opinions then serving as guidance to the district courts in the circuit. We have since explained that the presumption of validity is not weakened or destroyed where merely pertinent non-considered prior art is introduced, but that the offering party is more likely to carry the burden of persuasion imposed by 35 U.S.C. Sec. 282 when art more pertinent than that considered is introduced. SSIH Equipment S.A. v. ITC, 718 F.2d 365, 218 U.S.P.Q. (BNA) 678, 687 (Fed.Cir.1983); Stratoflex, supra, 218 USPQ at 875-76.

Citing Adams, supra, 383 U.S. at 48, 86 S. Ct. at 712, and Clark Equipment Co. v. Keller, 570 F.2d 778, 785 (8th Cir.1978), Judge Devitt stated: "Recognizing the presumption of validity, we must satisfy ourselves in light of the evidence before us that the patents in suit [sic patented inventions] possess the essential characteristics of utility, novelty and nonobviousness." It is of course true that to be patentable an invention must be useful, novel, and nonobvious. The cited cases do not hold, however, that a court may merely "recognize" the presumption and then proceed to "satisfy" itself that an invention possesses such characteristics. The statute, 35 U.S.C. Sec. 282 requires that the court begin by presuming that the invention has those characteristics. To do otherwise is to ignore the statute and to place on the patentee a non-statutory burden of proving validity. The court must be satisfied of something quite different, namely that the party challenging validity has carried its burden of overcoming the presumption. In this regard, this court has said that a court need not declare a patent valid, but need only in a proper case declare that the party challenging validity has not carried that burden. Environmental Designs, Ltd. v. Union Oil Co., 713 F.2d 693, 218 U.S.P.Q. (BNA) 865, 871, n. 9 (Fed.Cir.1983). As discussed, infra, the conclusion correctly reached on the entire record (that the asserted claims are invalid) carries with it a determination that CPI carried its burden of overcoming the presumption, and the misstatement concerning the role of the court did not constitute reversible error in this case.

On review of the record and independent analysis of the obviousness issue, infra, we conclude that none of the statements concerning the presumption of validity so influenced the appealed judgment as to require reversal.

The contention that an improper test was used for determining obviousness is based on the Memorandum statement that "obviousness of the invention is measured by the knowledge of a hypothetical person skilled in the art who has thought about the subject matter of the patented invention in light of that art " (emphasis ours). Medtronic says that statement indicates an improper reliance on the teachings of the patent in suit, i.e., that the test applied was one of hindsight. That approach would have been error. Orthopedic Equipment Co. v. United States, 702 F.2d 1005, 1012, 217 U.S.P.Q. (BNA) 193, 199 (Fed.Cir.1983); W.L. Gore & Associates, Inc. v. Garlock, Inc., 721 F.2d 1540 at 1552, (Fed.Cir., 1983). CPI, relying on conjecture, says the "subject matter of the patented invention" must have referred to the problem to be solved, not the solution set forth in the claims. The quoted expression, taken literally, is an improper statement of a test for obviousness, insofar as it rests on the notion that it is possible for those skilled in the art to think about an actual invention before the inventor makes it. The expression is at best ambiguous, particularly in light of another Memorandum statement in which hindsight was specifically eschewed. Absent other bases for the appealed judgment, it could not be said that a correct test was here applied. We review judgments, however, not statements in memoranda. Chore-Time Equipment, Inc. v. Cumberland Corp., 713 F.2d 774, 218 U.S.P.Q. (BNA) 673, 677 (Fed.Cir.1983). As appears infra, the judgment is so thoroughly supported in the record and in other portions of the Memorandum, as to preclude reversal on the basis that an improper test was actually applied.

Additional troubling language, not specifically identified by Medtronic, appears in the Memorandum statement that rate-limiting appears to be the "essence" of the Greatbatch invention. In determining obviousness/nonobviousness, an invention must be considered "as a whole", 35 U.S.C. Sec. 103, and claims must be considered in their entirety. Schenck, A.G. v. Nortron Corp., 713 F.2d 782, 218 U.S.P.Q. (BNA) 698, 700 (Fed.Cir.1983). Though applicable elsewhere, e.g., in determining infringement under the doctrine of equivalents, there is no legally recognized "essence" of the invention applicable in determining validity. Gore, supra, at 1548. Again, however, it cannot be said that the inaccuracy of the statement so influenced the appealed judgment as to require reversal.(2) The Merits Under 35 U.S.C. Sec. 103

The district court found the scope and content of the prior art, differences between the claimed invention and the prior art, and level of skill in the art for each patent, specifically following the guidance 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, 148 U.S.P.Q. (BNA) 459, 467 (1966).

(i) The Claimed Invention

NOTE: OPINION CONTAINS TABLE OR OTHER DATA THAT IS NOT VIEWABLE

Figure 1 of Greatbatch discloses a pacemaker 100 having an oscillator 10 producing pulses, amplified by an amplifying stage 12, for stimulating a heart at output terminals 16 and 18. The magnitude of the pacemaker output pulses is controlled by the voltage stored in capacitor 58, the larger the stored voltage the smaller the amplitude of each output pulse.

Runaway is inhibited by runaway inhibitor circuitry 20 having a switching transistor 72 which, when conductive, allows capacitor 58 to discharge. The conductivity of transistor 72 and hence the charge on capacitor 58 is controlled by the values of resistor 68 and capacitor 62, a resistor-capacitor (RC) timing circuit. Optimum operation occurs with the values of resistor 68 and capacitor 62 such that when oscillator 10 produces pulses at a rate of 60 pulses/min., transistor 72 is conductive for a sufficient time period to allow capacitor 58 to discharge completely.

As oscillator 10 produces pulses at a rate increasing above 60 pulses/min., transistor 72 is conductive for shorter time periods, capacitor 58 does not discharge completely, and the amplitude of each output pulse decreases. When oscillator 10 produces pulses at a rate of 120 pulses/min., transistor 72 is constantly nonconductive, capacitor 58 cannot discharge, and the amplitude of each output pulse decreases to zero. Thus, runaway is inhibited.

A second embodiment uses the switching transistor to disconnect oscillator 10 from the power supply 14 for 0.5 sec. after production of each pacemaker output pulse. With that fixed, built-in time delay, the pulse rate can never exceed 120 pulses/min.

The inhibitor circuitry 20 is characterized by the parties as being "independent" of oscillator 10, meaning that the output of the pacemaker is limited by the inhibitor circuitry regardless of the pulse rate of the oscillator. That independence is achieved by positioning the inhibitor circuitry 20 so that it will not be affected by runaway-inducing pacer malfunctions. The positioning is "downstream" of oscillator 10, i.e., between the output of oscillator 10 and output terminals 16, 18, as in the first embodiment, or between oscillator 10 and voltage source 14, as in the second embodiment. In either position, the inhibitor circuitry 20 prevents runaway of pacemaker 100 independently of any malfunctioning of oscillator 10 or other pacer malfunction.

Claims 1, 8, 10, 11, and 12 read on both embodiments. Claim 1 reads:

1. A cardiac Pacemaker which comprises a semiconductor pulse generator adapted to normally generate pulses at a first predetermined frequency, a power supply coupled to said pulse generator, a pair of Pacemaker output terminals coupled to said pulse generator and runaway inhibit means coupled to said pulse generator for preventing Pacemaker malfunction from causing pulses to be applied to said Pacemaker output terminals at a frequency rate above a second predetermined frequency higher than said first frequency.

Claims 8, 10, 11, and 12 all depend from claim 1. Claim 8 includes a timing circuit having a fixed time interval approximately equal to the period of the second predetermined frequency. Claim 10 includes a capacitor connected across the power supply. Claim 11 provides the cardiac pacemaker shall be implantable.

12. A cardiac Pacemaker according to claim 1 wherein said runaway inhibit means includes a switching transistor having a high impedance and a low impedance state, a capacitor, and a resistor connected in series with said capacitor and having a selected time constant, and circuit means connecting said capacitor and said resistor to said switching transistor and to said pulse generator for cyclicly controlling said switching transistor between high impedance and low impedance states.

In view of our disposition, we do not reach CPI's contention that Medtronic conceded invalidity of claims 10 and 11.

(ii) The Prior Art

The scope of the prior art includes pulse controlling circuitry within and without the cardiac pacemaker field. The content of that art is represented by (1) U.S. Patent 3,253,596 (Keller I), (2) U.S. Patent 3,299,295 (Goda), (3) U.S. Patent 3,187,202 (Case), and (4) advertisements in 1962 and 1964 (ads) for Medtronic's earlier 5860 and Electrodyne pacemakers, respectively. Of that art, only Keller I was cited by the PTO:NOTE: OPINION CONTAINS TABLE OR OTHER DATA THAT IS NOT VIEWABLE

Keller I discloses a pacemaker having an oscillator 40 responsive to natural atrial stimulation pulses of the heart to produce and deliver to the heart artificial ventrical stimulation pulses synchronized with the atrial pulses, i.e., a synchronous pacemaker. It further discloses a time delay or limiting circuit 30 interposed between amplifier input stages 10, 20 and the oscillator 40, i.e., "upstream" of the oscillator. Pulse generator 50 and output amplifier 60 comprise output means for delivering the output pulse to the heart. As a natural atrial input pulse is transmitted from the heart to the oscillator 40 through the time delay circuit 30, a capacitor C5 in the time delay circuit 30 charges and precludes the oscillator 40 from receiving further input pulses. The oscillator 40 remains isolated until capacitor C5 discharges through a resistor R5, the period of oscillator isolation (RC time constant) being determined by the values of capacitor C5 and resistor R5. Runaway caused by too many input pulses from the heart is thus inhibited but because circuit 30 is upstream of oscillator 40, it can be described as "dependent" on the oscillator's proper functioning insofar as circuit 30's ability to inhibit runaway of pacemaker output pulses is concerned.

Goda discloses a circuit 10 which inhibits runaway of pulses from a source:NOTE: OPINION CONTAINS TABLE OR OTHER DATA THAT IS NOT VIEWABLE

In Goda, pulses are input at input terminal 22 of input stage 15 and pulses are output at output terminal 23 of output stage 20. Between stages 15 and 20 is a pulse width control stage 25 which inhibits output of pulses having greater than a predetermined pulse width.

Also between stages 15 and 20 is a rate-limiting circuit 30 which inhibits runaway. A switching transistor 50, when conductive, essentially grounds point x, inhibiting output of pulses. The conductivity, i.e., high and low impedance states, of the switching transistor 50 is controlled by the values of resistors 82 and 84 and capacitor 78. Those values are such that when the input pulse rate exceeds a predetermined value, the capacitor 78 charges, causing switching transistor 50 to become conductive and inhibiting the output of pulses. The circuit 10 of Goda is downstream of the pulse source and, independently of any malfunctioning of that pulse source, circuit 10 inhibits runaway if pulses are produced at an unacceptably high rate.

Case also discloses a circuit which inhibits runaway of pulses from a pulse source. A capacitor charges in response to input pulses and discharges in their absence. When the input pulse rate exceeds a predetermined rate the capacitor does not have time to discharge. The voltage stored in the capacitor increases until the next stage is activated, which produces a feedback signal for inhibiting the output of pulses for a predetermined time. The circuit of Case is located downstream of the source of pulses and thus independently inhibits runaway.

The ads, though they contain no technical details, indicate in their relevant portions the advisability of employing circuitry to protect against pacemaker runaway caused by battery failure:

RUNAWAY PACEMAKER SAFEGUARD TR-14 circuitry automatically reduces pacemaker rate should battery failure occur ... thus protecting the patient against ill effects of too rapid a rate of stimulation.Circuitry in the new Model 5860 prevents the rate of the Pacemaker from ever increasing beyond a rate of 90--regardless if it is caused by battery expiration or the complete failure of an entire cell.

The difference between the claimed invention and the prior art found by Judge Devitt lies in the independence of the runaway inhibitor circuitry from the oscillator serving as the pulse source, in dependency of Keller I's upstream delay circuit 30 on the oscillator 40, and in Case and Goda's failure to specifically address the problem of a malfunction in the pulse source.

The unchallenged level of ordinary skill in the art was high, a minimum of a bachelor's degree in electrical engineering often being accompanied by a master's degree and several years of design experience.

Finding that the ads suggest use of circuitry to achieve runaway inhibitor independence, the district court concluded that it would have been obvious to apply the techniques of Case and Goda to the pacemaker of Keller I.

(iii) Medtronic's Contentions

Medtronic says several findings are clearly erroneous because: (1) Case and Goda teach control of pulse density, not pulse rate, and thus do not teach techniques for inhibiting runaway and are not more pertinent than the art considered by the PTO; (2) the ads do not suggest runaway inhibitor independence and are not more pertinent than the art considered by the PTO; (3) Keller (I) does not inhibit runaway in both synchronous and asynchronous modes; (4) runaway inhibitor independence is not the sole difference between the claimed invention and the prior art; (5) there is no suggestion in the prior art of how the teachings of the prior art could be combined to render obvious the claimed invention; and that Judge Devitt failed to consider facts constituting objective indicia of nonobviousness.

Contention (1) is not persuasive. A careful review of the references and the testimony concerning them makes plain that Case and Goda teach the control of pulse density and pulse rate.

The rate-limiting circuit of Case includes an integrator. The integrator allows the rate-limiting circuit to operate in a "burst mode" when a discrete number of high frequency pulses are not inhibited, in the expectation that the frequency will decrease (no runaway). After burst mode operation, if the frequency remains unacceptably high (runaway), the circuit operates to limit pulse rate. Thus, the record clearly establishes that the circuit of Case operates first in a burst mode and then as a rate-limiting circuit.

CPI's expert witness Prohofsky testified that Case's burst mode operation was an additional function added to the rate-limiting function of the circuit and that to one of ordinary skill it would have been obvious to employ the rate limiting function without the integrator when burst mode was not desired. Prohofsky testified in detail how Case would then operate to block, for example, every other pulse, clearly establishing rate-limiting rather than control of pulse density.

Medtronic complains that Prohofsky "simplified" the Case disclosure by replacing discrete components (resistors, capacitors, transistors, etc.) with boxes labeled to represent their appropriate functions (amplifiers, oscillators, etc.). That complaint is unavailing in this case. Showing a circuit as composed of labeled boxes, rather than as discrete components, can facilitate description and comprehension.

Simplification of a reference for description purposes cannot, however, be allowed to cross the line at which it becomes modification of that reference's disclosure. Here, Prohofsky's description of Case involved deletion of the integrator and elimination of burst mode from the operation of the overall circuit. Prohofsky's testimony describing the Case circuit when burst mode operation is not desired was not, however, an effort to show how Case could be modified in light of the Greatbatch invention. Nor was there here a picking and choosing from a reference, to the exclusion of material necessary to full appreciation of what the reference fairly suggests. In re Kamm, 452 F.2d 1052, 1056-57, 59 CCPA 753, 172 U.S.P.Q. (BNA) 298, 301-302 (1972). Prohofsky's total testimony, on direct and under cross-examination, was devoted to the question of whether it would have been obvious to those skilled in the art to employ the rate-limiting portion of the Case circuit in a pacemaker. We cannot, therefore, say that it was reversible error to have permitted him to describe the rate-limiting aspects of the Case disclosure in the manner reflected in the record.

No basis exists for a determination de novo on the present record that Case's burst mode is such an integral part of the Case rate-limiting circuitry that Case could not be said to suggest rate-limiting to those skilled in the art. Indeed, Case teaches:

A feature of this invention is a pulse count control circuit which is responsive to a source of pulses and includes means to sample the pulses and to set a reference pulse count rate level. If the reference pulse count rate level is exceeded, a control pulse is generated which initiates a gate and this gate permits a control pulse to inhibit the pulse input during a predetermined time interval. (emphasis added).

Prohofsky's testimony having been credited at trial, and the high level of skill in the art being unchallenged, we cannot say that Medtronic has met its appellate burden of showing that the finding that Case teaches rate-limiting was clearly erroneous.

Much of what has been said concerning Case is equally applicable to Goda. Goda does not disclose a separate integrator but relies on a rate-limiting circuit to perform integration. Nor does Goda disclose values for circuit components, relying on the user to choose values compatible with desired circuit operation. Prohofsky testified that those skilled in the art would have considered it obvious to choose component values such that the number of pulses in the burst mode could be as low as two, and, when burst mode is not desired, to choose values such that no integration is performed.

[S]tage 30 senses the rate or duty cycle of the input pulses, in order to control the rate of the output pulses, by inhibiting the production of such pulses whenever the rate of the input pulses exceeds a given limit. This limit is controlled by selecting the integrating network of the stage 30 to switch transistor 50 to a conducting state whenever the average potential of the collector 40c equals a given value. Thus, stage 30 may be thought of as a pulse rate control stage. (emphasis added).

For the reasons applicable to Case, it cannot be said that Medtronic has met its appellate burden of showing that the finding on Goda was clearly erroneous.

We agree with the district court that Case and Goda are more relevant in a major respect than Keller I. In the pacemaker circuit of Keller I, the ability of the time delay circuit 30 to control final output is dependent on the oscillator. In the circuits of Case and Goda, like that of Greatbatch, the ability of the rate-limiting circuits to control final output are independent of the oscillator.

Prohofsky compared Fig. 1 of Greatbatch to Fig. 1 of Goda, reflecting a close correspondence: Goda's transistor 50 corresponding to Greatbatch's transistor 72; Goda's resistors 82 and 84 corresponding generally to Greatbatch's resistor 68; Goda's capacitor 78 corresponding to Greatbatch's capacitor 62. Prohofsky further testified that "[t]he Greatbatch circuit accomplishes its inhibition by turning transistor 74 [sic 72] off when the rate is exceeded. The Goda circuit accomplishes inhibition by turning transistor 50 on". The credited testimony establishing that the correlation between the Goda and Greatbatch circuits is closer than that between the Keller I and Greatbatch circuits in respect of rate-limiting cannot be simply ignored on appeal.

Medtronic's contention that a pacemaker designer in 1964 would not have looked to Case or Goda, solely because those patents disclose circuits used in high power, high frequency devices, is not persuasive. Faced with a rate-limiting problem, one of ordinary skill in the art would look to the solutions of others faced with rate-limiting problems. Weathering Engineering Corp. of America v. United States, 204 U.S.P.Q. (BNA) 41 (Ct.Cl. tr. div. 1979), aff'd, 208 U.S.P.Q. (BNA) 939 (1980). Case and Goda deal with the problem of rate limiting.

The ads clearly disclose the benefits of, and thus suggest, circuitry to inhibit runaway in a pacemaker and for that reason are pertinent prior art. Because the ads do not suggest that the circuitry should be independent (downstream) of the oscillator, the finding that they do is clearly erroneous. The pertinency of the ads describing Greatbatch's prior pacemaker, however, is made clear by the reference in the Greatbatch patent to particular modes of failure and by the specific mention in the patent that Greatbatch's prior pacemaker "is fail-safe from normal battery exhaustion". The ads deal with battery failure including expiration and also indicate that the prior pacemakers are fail-safe from failure of an entire cell. The 5860 ad says its circuit prevents the rate from ever increasing above 90. The ads thus deal with a larger group of problems beyond mere battery exhaustion. The finding on pertinency of the ads is not clearly erroneous.

Medtronic's attack on the finding that Keller I provided runaway inhibition in the synchronous and asynchronous modes must fail. Prohofsky's testimony concerning circumstances in which the Keller I circuit performs some runaway inhibiting in both modes provides fully adequate support for that finding, and the presence of contrary testimony does not render it clearly erroneous.

Medtronic says characterization of the difference between the claimed invention and the prior art as solely "runaway inhibitor independence" is erroneous because no prior art suggested use of a circuit to prevent a pacemaker from stimulating the heart above a predetermined safe rate when a pacemaker malfunctions. "Pacemaker malfunction", however, includes battery exhaustion, and the ads clearly indicate that circuitry can prevent a pacemaker stimulating the heart beyond a predetermined safe rate, whether the pacemaker malfunction is caused by battery exhaustion or by failure of an entire cell. Thus Medtronic has not shown the finding on the difference between the claimed invention and the prior art, including the ads, to have been clearly erroneous.

Medtronic says there is nothing in the prior art to suggest combining its teachings as done in Greatbatch. As above indicated, the analysis in the Memorandum is flawed by the erroneous finding that the ads suggest locating the runaway inhibiting circuitry downstream of the oscillator pulse source. That erroneous finding is of no moment, however, in light of the disclosure of that location in Case and Goda and the consequent support in the record for the conclusion that it would have been obvious to employ the downstream circuit of either Case or Goda in the pacemaker of Keller I.

The ads clearly disclose the advisability of runaway inhibitor circuitry to protect against pacemaker runaway. One skilled in the art faced with the problem of preventing runaway, with that suggestion to use circuitry, would look for a solution among circuits employed by others faced with the same problem. The solution taught by Case and Goda, i.e., location of runaway inhibitor circuit elements downstream or independently of the pulse source, is precisely that disclosed and claimed in the Greatbatch patent.

That one skilled in the art, considering the teachings of the references as a whole, would have found it obvious to apply the runaway inhibiting circuitry of Case and Goda to a pacemaker is amply demonstrated in the record. Prohofsky testified without contradiction that the rate-limiting circuit 30 of Goda:

is really a very simple sort of thing. This is what one normally does in developing RC timing circuits--is to establish a time arrangement through a resistor-capacitor configuration and sensing some condition on that capacitor with a transistor.

Where the prior art teaches the advisability of runaway inhibitor circuitry in pacemakers (ads), use of RC timing circuits downstream of a pulse source (Case and Goda), and use of RC time delay circuits upstream of an oscillator in a pacemaker (Keller I), and the record reflects that those skilled in the art normally use RC circuits to establish timing arrangements (Prohofsky's testimony), it would have been obvious when the Greatbatch invention was made to provide a pacemaker with an RC timing circuit downstream of the oscillator pulse source.

None of the limitations expressed in the other asserted claims leads to a different conclusion. We agree with the district court's determination that CPI carried its burden of proving that the inventions not only of claim 1 but of claims 8, 10, 11, and 12 as well, would have been obvious.

When objective evidence of nonobviousness is available it must be considered. In re Sernaker, 702 F.2d 989, 996, 217 U.S.P.Q. (BNA) 1, 7 (Fed.Cir.1983). The parties stipulated to exclusion of evidence of commercial success. Though Medtronic suggests the filling of a long felt need, it introduced no evidence that the industry so recognized the Greatbatch invention. Wilson Greatbatch, the inventor and individual in charge of design control at Medtronic from 1960 to 1970, testified that Medtronic itself did not for many years use the pacemaker of Greatbatch. There are thus no objective indicia of nonobviousness available for weighing on the scale in deciding the obviousness issue.

Considering what the references as a whole would suggest to one of ordinary skill in the art, recognizing that Case, Goda, and the ads are more relevant than the references considered by the PTO, and noting the lack of objective indicia of nonobviousness, we will affirm the judgment based on Judge Devitt's determination that CPI carried its burden of proving that the inventions set forth in the asserted claims of the Greatbatch patent would have been obvious at the time those inventions were made. The erroneous finding that the ads suggest locating the inhibitor circuitry downstream of the oscillator does not, on this entire record, require a contrary result.

(i) The Claimed InventionNOTE: OPINION CONTAINS TABLE OR OTHER DATA THAT IS NOT VIEWABLE

Wingrove discloses a pacemaker having a set counter 40 storing a number. A clock counter 70 counts pulses produced by a clock 80. Comparison logic 50, responsive to set counter 40 and clock counter 70, produces a pulse when the number of clock pulses counted by the clock counter equals the number stored in the set counter. That output pulse triggers a one shot multivibrator 90, producing an output pulse available through an output pulse circuit 100 at output terminals 108 and 109 to stimulate a heart. The output pulse of the multivibrator 90 resets clock counter 70, allowing the process to repeat.

After the pacemaker has been implanted, the number stored in set counter 40 may be reprogrammed by transmitting from outside the bearer's body an encoded radio frequency signal representative of the desired number. The transmitted signal is received by receiver/filter 10 and input to decoder 20 which prevents spurious signals from affecting the number stored in set counter 40. Decoder 20 inputs the new number into set counter 40 and the pacemaker continues operating as discussed above.

The pacemaker of Wingrove is described in the Memorandum as employing digital two-counter architecture to achieve programmable frequency division: "two-counter architecture" because of set counter 40 and clock counter 70; "programmable" because the number stored in set counter 40 may be changed; and "frequency division" because the rate of pulse output by clock 80 is divided by the number stored in set counter 40.

1. Electro-medical stimulation apparatus comprising: first and second binary data storage means; means for selectively changing data stored in the first storage means; means for continually changing data stored in the second storage means at a predetermined rate; means for comparing at least portions of data stored in the first and second storage means and for providing signals in the presence of a predetermined comparisons; and means connected to receive the signals from the means for comparing for providing stimulation signals.

Claim 4, dependent on claim 1, includes receiver and decoder means. Claim 5, dependent on claim 4, includes transmitter and encoder means.

Apparatus for providing electro-medical stimulation pulses comprising: clock means having a pulsed output signal; first counter means connected to count the output of the clock means; further counter means for receiving and storing a predetermined count; comparison logic means for continually comparing the counts in the first and further counter means to provide a polarity [sic] of signals; and stimulation circuit means controlled by the plurality signals for providing stimulation signals.

Claim 8, dependent on claim 7, includes means for receiving pulse signals. Claim 9, dependent on claim 8, includes means for transmitting pulse signals and means for varying the number of pulses transmitted. Claim 10, dependent on claim 9, includes encoding and decoding means.

23. In implantable cardiac pacer apparatus, including electrode means adapted to be connected to a heart and circuit means for providing electrical impulses on the electrode means, the improvement comprising: electrical storage means for storing programs and having at least first and second storage portions; means for receiving remotely sent programs connected to the first storage portion; clock means; means connecting the clock means to the second storage portion; means for comparing programs stored in the first and second storage portions for providing signals; means for connecting the means for comparing between the first and second storage portions and the circuit means, for controlling the electrical impulses according to the stored program; and all said means encapulated [sic] in means substantially inert to body fluids and tissue.

(ii) The Prior Art

Making the factual inquiries of Graham, supra, Judge Devitt found the scope of the prior art to include electronics generally, digital circuitry as used in pacing devices, including digital two-counter architecture, and remote radio control. The content of that art is represented by (1) U.S. Patent 3,557,796 (Keller II), (2) U.S. Patent 3,537,003 (Planta), (3) U.S. Patent 3,662,758 (Glover), (4) U.S. Patent 3,631,860 (Lopin), (5) Walsh & Moore, Digital Timing Unit for Programming Biological Stimulators, American Journal of Medical Electronics, 29-34 (First Quarter 1966) (Walsh & Moore), and (6) U.S. Patent 3,629,710 (Durland). Of that art, Keller II, Glover, Lopin, and Durland were cited by the PTO.

Keller II discloses a pacemaker that controls the production of output pulses with a digital counter driven by an oscillator. Its significance lies in its recognition of the advantages derivable from use of miniaturized integrated digital circuits, including a digital clock time base, in an implantable cardiac pacemaker.

Planta discloses a frequency measuring apparatus which includes a measuring or set counter for storing a number, a clock counter for counting pulses produced by a clock, and comparison logic for producing an output signal when the number of pulses counted equals the number stored. The number stored in the set counter can be reprogrammed. Planta therefore teaches digital, two-counter architecture which achieves programmable frequency division. In terms of the claim language Planta teaches first and second binary data storage means, means for selectively changing data stored in the first storage means, means for continually changing data stored in the second storage means at a predetermined rate, and means for comparing at least portions of data stored in the first and second storage means and for providing signals in the presence of predetermined comparisons.

Glover discloses an implantable stimulator adapted for attachment to a muscular organ such as a bladder. Radio frequency signals transmitted from outside the bearer's body supply power and control signals to the stimulator.

Lopin discloses a variable rate pacemaker using a resistor-capacitor time constant (analog approach) to control the output of pulses. A counter defines four possible states. Each state determines a different magnitude of charging current which determines the rate of the pacemaker. Magnetic reed switches are used to cycle the counter until the desired one of the four possible states is reached, thereby establishing the desired rate. After implantation, the pacing rate can be changed by placing a magnetic field in the vicinity of the patient's chest to change the positions of the magnetic switches, cycling the counter until the new rate is established.

Walsh & Moore discloses a timing unit including a series of digital counting modules connected to form a timing chain. Each module counts output pulses from the previous module, with the first module counting output pulses from a high frequency clock. Each module outputs a signal for each ten input signals. Ten other output terminals available at each module indicate how many pulses, from zero to nine, have been input to the module and are each controlled by a switch. Counting continues until an output pulse representing the manually selected time interval (determined by switch positions) is produced and resets the modules to start counting anew.

Durland discloses a pulse generator using digital, two-counter architecture to control the ratio of the pulse width to the pulse interval. A recirculating counter is driven by 60 cycle line frequency. A coincidence detector determines when the value in the recirculating counter equals the variable value stored in a data register.

Unchallenged on appeal is the finding that the components of digital two-counter architecture (a set counter, a clock, a clock counter for counting clock pulses, and a comparator) were available from RCA and others as ready-to-use, integrated, packaged circuits.

The district court found that the difference between the claimed invention and the prior art was use of digital two-counter architecture to control rate of pulse generation in a pacemaker, and that the level of ordinary skill in the art was high, with bachelor's and master's degrees in electrical engineering common.

In light of the teachings of Keller II, Planta, Walsh & Moore, and Lopin, the district court correctly concluded that it would have been obvious to use digital two-counter architecture to achieve programmable frequency division in an implantable pacemaker. The record makes clear that use of radio frequency signals to control an implanted device would have been obvious from Glover and that encoding/decoding of signals would have been obvious from U.S. Patent 3,445,815 to Saltzberg and Terry and a 1969 article entitled "Basics: Remote Radio Controls", both introduced at trial and disclosing use of encoding transmitters and decoding receivers to guard against false operation from noise or interference.

(iii) Medtronic's Contentions

Medtronic says there are clearly erroneous findings, because: Planta and Walsh & Moore are not more relevant than art considered by the PTO; nothing in the art would have suggested Wingrove's claimed invention; and objective indicia of nonobviousness were not considered.

As discussed above, Planta discloses digital two-counter architecture and programmable frequency division. That the latter is used in Planta for purposes other than timing output pulse rate does not diminish its relevance. Though Durland is relevant, Planta is more so. Planta's clock counter, like that of Wingrove, counts pulses produced by a clock element of the circuit, whereas Durland's counter is responsive to 60-cycle line frequency, a source not available in an implantable pacemaker. Further like Wingrove, Planta discloses a set counter to store the desired number whereas Durland discloses bistable elements (flip-flops or magnetic core elements) as storage means.

Walsh & Moore discloses a digital timing unit. Keller II discloses a digital timer in a pacemaker, using hardwired outputs of a counter rather than switches as in Walsh & Moore. Thus Walsh & Moore add nothing to what was before the PTO in Keller II. Though it was therefore error to consider Walsh & Moore more relevant than the art considered by the PTO, the error is harmless on the entire record and is insufficient to require reversal of the appealed judgment in light of other prior art teachings not considered by the PTO and effective to render obvious Wingrove's claimed invention.

Keller II discloses the advantages of using digital circuitry in a pacemaker:

An advantage of the use of digital circuitry for providing the various timing functions is that the various components, e.g., the counters, gates and flip-flops, can be constructed in integrated circuit form using presently available devices. Further, by custom designing the integrated circuitry using presently available fabrication techniques, the entire pacer can be constructed on a single semiconductor substrate thereby further reducing size, eliminating hand-wired circuit interconnections and facilitating hermetic sealing. Further, such integrated circuits have relatively low power consumptions.

With Keller II's advocacy of digital circuitry in pacemakers, it would have been obvious to one of ordinary skill in the art to select Planta's two-counter architecture (available in a ready-to-use integrated circuit), to achieve programmable frequency division in a pacemaker, and to use encoded/decoded radio signals as demonstrated by Glover and the art introduced at trial to achieve remote programmability.

Medtronic's objective indicia of nonobviousness argument rests on a statement of CPI's chairman that Wingrove's pacemaker was "quite sophisticated", expert testimony that Wingrove's invention opened the door for modern programmable pacemakers, and that most pacemakers made today are of the Wingrove type. Sophistication, however, is not alone synonymous with nonobviousness, particularly in a crowded, technically sophisticated art, where the ordinary skill was high. That the Wingrove invention opened the door for modern pacemakers and that most of the pacemakers made today are of the Wingrove "type" is not evidence that the claimed invention was recognized by the industry. The objective evidence falls short of establishing that Wingrove's claimed invention as a whole would have been nonobvious in light of art more relevant than that considered by the PTO.

Considering what the references as a whole would suggest to one of ordinary skill in the art, that Planta is more relevant than the art considered by the PTO, and that the showing of objective indicia of nonobviousness was non-probative, we will affirm the determination that CPI carried its burden of proving invalidity of claims 1, 4, 5, 7, 8, 9, 10, and 23 of Wingrove.

(i) The Claimed Invention

NOTE: OPINION CONTAINS TABLE OR OTHER DATA THAT IS NOT VIEWABLEWalmsley discloses an electromedical pulse generator having a power source 1, sensing circuit 2, oscillator circuit 3, and pulse generator 4 producing output pulses available at terminals 5 and 6. Sensing circuit 2 senses a drop in the output voltage of power source 1 below a predetermined level. Oscillator 3 is responsive to sensing circuit 2 and is connected in controlling relation to pulse generator 4 for automatically decreasing the rate at which pulse generator 4 produces output pulses.

13. In an electromedical pulse generator used to stimulate a selected portion of the body of the type having terminal means adapted for connection to a body implantable electrical lead, having power source means and having electrical impulse generating means electrically connected to the power source means for supplying output pulses to the terminal means at a first predetermined rate, the improvement comprising: means for sensing a significant decrease in the power source output below a predetermined level, means responsive to the sensing means for fixing the output pulse rate at a second diverse predetermined rate over a predetermined range of power source outputs, means connecting the sensing means to the power source means, and means connecting the responsive means to the sensing means and the impulse generating means.

(ii) The Prior Art

The district court correctly found that the scope of the prior art included methods of sensing loss of battery voltage within and without the field of cardiac pacemaker design. The content of that art is represented by U.S. Patent 3,503,062 (Witzke), Thornander, Asynchronous Pacers as Part of a Block-Building System, 167 Annals of the New York Academy of Sciences, 858 (1969), neither of which was considered by the PTO, and prior art references considered by the PTO that teach gradual rate change as an indication of battery depletion.

Witzke discloses a power supply voltage level indicator:

NOTE: OPINION CONTAINS TABLE OR OTHER DATA THAT IS NOT VIEWABLE

A device 10 is supplied with power from a battery 12. The voltage level indicator 14 is comprised of a voltage level detector 16 selectively controlling oscillators 18, 20 in response to the output voltage level of the battery 12. Oscillator 18 operates at a first rate (f1 ) that is different from a second rate (f2 ) at which the oscillator 20 operates. Oscillators 18, 20 energize a neon light bulb 22 through a driver 24. When the output voltage level of battery 12 is satisfactory, oscillator 20 causes bulb 22 to flash at the second rate f2 . If the output voltage level drops to a predetermined unsatisfactorily low value, oscillator 18 causes bulb 22 to flash at the first rate f1 .

Thornander discloses that in pacemakers using oscillators of the blocking type, pulse rate rises when battery voltage drops. In oscillators of the multivibrator type, pulse rate remains stable when battery voltage drops. Stability is a disadvantage in a pacemaker because there is no simple way of estimating when battery voltage is critically dropping. To overcome that disadvantage, "deterioration" is built into the rate-stable circuit by addition of a nonlinear element in the resistance chain that determines rate. The nonlinear element can be chosen to provide, for example, a drop from 70 to 60 pulses per minute in response to a drop in voltage caused by loss of a single battery cell thus providing an indication of battery exhaustion or cell failure.

The unchallenged level of skill in the art was high, a bachelor's degree in electrical engineering with digital circuitry design experience being common.

The difference between the claimed invention and the prior art was found to lie in use of step, (as opposed to continuous) rate change as an indication of battery depletion.

The PTO originally rejected Walmsley's application because U.S. Patent 3,486,071 (Hedge) disclosed a circuit for continuously monitoring battery voltage, the time for charging an inductor increasing as battery voltage decreased. Walmsley amended his claim to distinguish from the continuously responsive means of Hedge, inserting a limitation to a voltage sensing means responsive only to a significant, predetermined voltage decrease. Witzke discloses a voltage sensing circuit that responds to a significant, predetermined voltage drop.

(iii) Medtronic's Contentions

Medtronic says there were clearly erroneous findings: (1) Witzke is no more pertinent than U.S. Patent 3,349,386 (Zug) considered by the PTO; (2) there is no suggestion in the art of Walmsley's invention; (3) Thornander is no more pertinent than U.S. Patent 3,747,353 (Keller III) considered by the PTO; (4) error occurred in the Memorandum statement that "one might contemplate the use of a step-like change of rate simply because of the discrete-state nature of digital circuits"; and that there was a failure to consider objective indicia of nonobviousness.

Zug discloses a battery discharge indicator having an indicator light energizing switch responsive to two trigger circuits which alternately render the switch conductive and nonconductive and thus cause the light to blink in response to a battery voltage decrease to a predetermined level. Medtronic correctly asserts that Zug and Witzke disclose circuits responsive to predetermined voltage drops and sensing circuits that do not change the operative output of the device whose power supply is being monitored. However, Zug does not disclose means for producing pulses at a first predetermined rate, as does Witzke (Oscillator 20), nor does Zug disclose means for fixing the output pulse rate at a second diverse rate in response to a voltage drop as does Witzke (Oscillator 18). The non-cited Witzke is thus more pertinent than Zug.

Medtronic's contention that there is nothing in the prior art to suggest Walmsley's invention founders on the presence of Witzke. Level detector 16 of Witzke performs the same function as Walmsley's claimed sensing means and is connected to the power source as is Walmsley's claimed connecting means. Witzke's oscillator 18 fixes an output pulse rate at a second diverse predetermined rate and is connected to the level detector 16. That Witzke's oscillator 18 output is not that of his device 10 and that oscillator 18 is not connected to device 10 does not remove from Witzke its ample suggestion of Walmsley's claimed invention.

Keller III, cited by the PTO, discloses an implantable pacemaker using a multivibrator as an oscillator. Like Thornander, Keller III recognized that pulse rate remains stable even though battery voltage drops. Like Thornander, Keller III places a nonlinear element (zener diode) in the resistance chain that determines rate. Thornander thus adds nothing to the teachings of Keller III, and the finding that Thornander is more pertinent is therefore clearly erroneous. In light of the Witzke reference, however, the error is harmless and cannot justify reversal of the appealed judgment.

Similarly, it was error to conclude that "one might contemplate" use of a step change simply because digital circuits are discrete. The test under Sec. 103 is not what "one might contemplate". The proper test is whether the references, taken as a whole, would suggest the invention to one of ordinary skill in the art. In re McLaughlin, 443 F.2d 1392, 1395, 58 CCPA 1310, 170 U.S.P.Q. (BNA) 209, 212 (1971). The district court did not, however, reach its ultimate conclusion solely on a "might contemplate" approach. Also considered and discussed were Witzke, Thornander, and the amendment to overcome Hedge. On the basis of all the evidence, not merely the discrete nature of digital circuits, the district court concluded that the invention of claim 13 would have been obvious. We cannot say on this record that the reference to "might contemplate" so controlled or influenced that conclusion as to require reversal.

Medtronic lastly contends that objective indicia, including satisfaction of a long felt need and CPI's filing of a patent application directed to a step change in output rate, were not considered. There is no warrant for assuming that the assertion of those indicia was not considered. Failure to consider any evidence, as above indicated, would have been error. However, the record does not support the assertion that Walmsley's invention satisfied a long felt need, and Medtronic did not prove that CPI knew of Witzke when it filed an application, abandoned before any prior art rejection, on a device similar to that claimed in Walmsley. Further, if it be assumed that CPI once viewed an invention like Walmsley's nonobvious, that assumption could not foreclose a court from reaching a contrary conclusion in light of the entire record. See Paramount Publix Corp. v. American Tri-Ergon Corp., 294 U.S. 464, 477, 55 S. Ct. 449, 455, 79 L. Ed. 997 (1935). On this record, consideration of the asserted indicia does not require reversal of the conclusion respecting obviousness.

The Jepson format of claim 13 (and claim 23 of Wingrove) was disregarded at trial and on appeal. When a claim of that type is asserted, a determination of whether the implied admission, that the preamble must be considered as prior art, should form a part of the analysis. In re Ehrreich, 590 F.2d 902, 909-10, 200 U.S.P.Q. (BNA) 504, 510 (CCPA 1979); In re Fout, 675 F.2d 297, 213 U.S.P.Q. (BNA) 532 (CCPA 1982). Because that question was not presented, and because the claimed invention as argued would have been obvious at the time it was made to those skilled in the art, we need not and do not decide the question here.

Considering what the references as a whole would suggest to one of ordinary skill in the art, that Witzke is more relevant than the references considered by the PTO, and that objective evidence of nonobviousness is lacking, we will affirm the conclusion that the invention set forth in claim 13 of Walmsley would have been obvious under Sec. 103.

The Memorandum contains a statement that the Walmsley "patent" would have been obvious. Claimed inventions, not patents, form the subject of the obviousness/nonobviousness determination. Only claim 13 was asserted. No evidence touching the validity of other claims in the Walmsley patent was presented. The statute, 35 U.S.C. Sec. 282, requires that claims be considered independently of each other. Insofar as the appealed judgment relates to such other claims, therefore, it must be modified to restrict it to claim 13 and then affirmed as modified.

(3) Infringement

The finding that the claims in suit are not infringed apparently flowed from the conclusion on invalidity. It does not rest on a described application of the asserted claims to the accused devices. Where, as here, the invalidity holdings are upheld, failure to resolve the infringement issue at trial is of limited effect. Had the invalidity determinations been reversed in relation to one or more of the three patents in suit, however, remand for further proceedings would have been required, placing at risk the potential for additional delay and expense, the possibility of recalling witnesses, a second appeal, and a consequent inefficient use of judicial resources. Stratoflex Inc. v. Aeroquip Corp., 713 F.2d 1530, 218 U.S.P.Q. (BNA) 871, 880 (Fed.Cir.1983). Further, evidence of infringement may bear on resolution of other issues. For example, "[a]n alleged infringer's lauding of all the available prior art may ... have a hollow ring when played against its disregard of that art and its copying of the invention." Stratoflex, supra, at 880. Evidence of copying may defeat assertions of nonenablement, Eibel Co. v. Paper Co., 261 U.S. 45, 65-66, 43 S. Ct. 322, 329, 67 L. Ed. 523 (1923), or nonutility. See, E.I. DuPont De Nemours & Co. v. Berkley & Co., Inc., 620 F.2d 1247, 1258-59, 205 U.S.P.Q. (BNA) 1, 9 (8th Cir.1980). The better practice is to decide both issues when both are presented at trial.

Though an invalid claim cannot give rise to liability for infringement, whether it is infringed is an entirely separate question capable of determination without regard to its validity. Because both validity and infringement involve construction of a claim, and because the construction must be the same in determining both, it is desirable to decide both questions at the same time.

We affirm the judgment that the inventions set forth in claims 1, 8, 10, 11, and 12 of Greatbatch, in claims 1, 4, 5, 7, 8, 9, 10, and 23 of Wingrove, and in claim 13 of Walmsley would have been obvious at the time those inventions were made to one skilled in the art and that those claims are therefore invalid under 35 U.S.C. Sec. 103. We modify the judgment by restricting its application to the foregoing claims.

MODIFIED AND AFFIRMED.