This is an appeal from a final judgment of the United States District Court for the Eastern District of New York in a patent infringement suit. CVI/Beta Ventures v. Tura LP, 905 F.Supp. 1171 (E.D.N.Y.1995). CVI/Beta Ventures, Inc. (CVI) is the assignee of United States Patent Nos. 4,772,112 (the ’112 patent) and 4,896,955 (the ’955 patent). The patents relate to flexible eyeglass frames. Marchon Eyewear, Inc. (Marchon) is the exclusive licensee under the patents. Marcolin USA, Inc. (Marcolin) and Rothandberg, Inc. (Rothandberg) are corporations which are affiliated with Marchon. We refer to CVI, Marchon, Marcolin, and Rothandberg collectively as “CVl/Marchon.” Tura LP and Brodart Co. (collectively, “Tura”) supply eyeglass frames to eyeglass retailers and distributors.
CVI and Marchon sued Tura in the United States District Court for the Eastern District of Virginia, alleging that various Tura frame models infringed the ’112 and ’955 patents. After the ease was transferred to the Eastern District of New York, Marcolin and Rothandberg were joined as plaintiffs. In its answer, in addition to denying infringement, Tura asserted counterclaims under the antitrust laws and a counterclaim for patent misuse under 35 U.S.C. § 271(d). Tura also sought a declaratory judgment of patent invalidity. Following a jury trial, claims 1-3, 5, and 6 of each patent were found to be not invalid and to be infringed by Tura LP, while Brodart Co. was found hable for inducing infringement. CVI/Beta Ventures, 905 F.Supp. at 1175. The jury further found that CVI was entitled to a reasonable royalty of 6%, and that Marchon, Marcolin, and Rothandberg collectively were entitled to lost profits and price erosion damages. Id. at 1204. In view of its verdict, the jury did not reach any of Tura’s counterclaims.
Following the denial of Tura’s renewed motion for a judgment of non-infringement as a matter of law (JMOL), judgment was entered in favor of CVI against Tura in the total amount of $819,208, plus prejudgment interest and attorneys fees, and in favor of Marchon, Marcolin, and Rothandberg against Tura in the total amount of $12,585,093, plus prejudgment interest. Id. In addition, Tura was enjoined from further infringement.1 Id.
Tura now appeals the final judgment of the district court. It argues that the court erred *1149in its claim construction in deciding the renewed JMOL motion and that it erroneously instructed the jury to presume that one of the limitations of claim 5 of the patents was found in certain of Tura’s accused frames. It also argues that Marcolin and Rothandberg were improperly joined as plaintiffs because they lack standing.2
We conclude that the district court did err in its claim construction. We also conclude that, when the claims are properly construed, because of CVI/Marehon’s failure of proof, Tura is entitled to a judgment of noninfringement as a matter of law. Accordingly, we (i) reverse the judgment of infringement, (ii) vacate the award of damages and the injunction entered against Tura, and (in) remand the case to the district court for entry of judgment of noninfringement in favor of Tura and for further proceedings in connection with Tura’s counterclaims. In view of our decision, we do not reach the jury instruction issue or the issue of Marcolin and Rothandberg’s standing.
BACKGROUND
I.
The ’112 patent was issued on September 20, 1988. The ’955 patent was issued on January 80, 1990, as a continuation-in-part of the application that resulted in the ’112 patent. The named inventors on the two patents are Robert B. Zider and John F. Krumme. On May 21, 1991, a reexamination certificate was issued for the ’955 patent, confirming the patentability of all of the claims of the patent. It is stated in each of the patents that the purpose of the invention is “to provide eyeglass frames which (1) are highly resistant to permanent deformation, or ‘kinking,’ over the full range of ambient temperatures, or (2) are sufficiently resistant to deformation and are readily restorable to the undeformed shape by heating ....” ’112, col. 2, lines 21-26; ’955, col. 2, lines 21-26.
The eyeglass frames invented by Zider and Krumme are fabricated from a “niekel-titanium based shape-memory alloy.” A shape-memory alloy is an alloy which exhibits the ability to return to its original shape after being deformed. In other words, the alloy “remembers” its original shape and seeks to return to that shape after a deforming force is removed.
The prior art discussed shape-memory alloys as potentially useful in eyeglass frames. There was, however, an obstacle to effective use of such alloys in frames. The obstacle lay in the fact that the shape-memory characteristics of the alloys only exhibited themselves over a narrow temperature range. As stated in the ’955 patent, that range was “too narrow to be of service to eyeglass frames which must function in winter days as cold as -20°C. and in hot sunny days with possible temperatures over 40°C.”3 ’955, col. 2, lines 8-10. Zider and Krumme overcame this obstacle by inventing eyeglass frames fabricated from a nickel-titanium based shape-memory alloy that had been modified by work-hardening (hammering, pressing, or stretching the alloy) or by a combination of work-hardening and heat treatment. ’112, col. 2, lines 29-51; ’955, col. 2, lines 21-55. As a result, Zider and Krumme’s eyeglass frames have the ability to resist deformation over an ambient temperature range of -20°C. to +40°C.
Both the ’112 and the ’955 patents have eight claims, two of which (numbers 1 and 5) are independent. The claims of the patents are substantially similar. Claims 1 and 5 of the ’955 patent are illustrative:
1. An eyeglass frame having at least a portion thereof fabricated from nickel-titanium based shape-memory alloy, said portion being in the work-hardened pseudoelastic metallurgical state, said portion having been subjected to work-hardening and having a low effective elastic modulus giving a soft, springy feel, said portion having greater than 3% elasticity over a temperature range from -20°C. to +40°C.
•b
5. An eyeglass frame having at least a portion thereof fabricated from nickel-tita*1150nium based shape-memory alloy, said portion being in the work-hardened and heat-treated condition, said portion having been subjected to at least 30% work-hardening followed by a heat-treatment at a temperature not exceeding 400°C. for not less than one hour and having a minimum of 3% heat-recoverable shape-memory, a yield strength greater than 30,000 psi and at least 3% elasticity.
’955, col. 13, line 33 — col. 14, line 5, col. 14, lines 16-24.
Claim 1 of the ’112 patent differs from claim 1 of the ’955 patent in that (i) it does not contain the word “pseudoelastic,” (ii) it specifies “at least 30%” work-hardening, and (iii) it refers to “4%” elasticity. ’112, col. 12, lines 14-22. The only difference between claim 5 of the ’112 patent and claim 5 of the ’955 patent is that the former calls for “20%” work-hardening. ’112, col. 12, lines 33-41. In both patents, dependent claims 2-4 and 6-8 relate to eyeglass temples, bridges, and nosepads. As will be seen, the claim construction issue in this case centers on the 3% and 4% elasticity limitations.
As used in the patents, the term “pseudoelastic” refers to the ability of a shape-memory alloy to return to its original shape after deformation, within a narrow ambient temperature range. Thus, in claim 1, the inventors claim eyeglass frames fabricated from nickel-titanium based alloys which have “pseudoelastic” shape-memory characteristics but which have been work-hardened so that they resist permanent deformation to the extent indicated (“greater than 3% elasticity”) over a useful ambient temperature range. The eyeglass frames claimed in claim 5 of the patents also are fabricated from nickel-titanium based shape-memory alloys which have been work-hardened. In addition, however, the alloys have undergone heat-treatment. Heat-treatment increases the ability of the alloys to resist permanent deformation by imparting to them the quality of “heat-recoverable shape-memory.” A component made from an alloy which exhibits “heat-recoverable shape-memory,” if stressed by a deforming force at a temperature below the alloy’s transformation temperature,4 may exhibit some amount of residual strain when the stress on it is released. This strain is recovered, however, and the component returns to its original shape, when the component is warmed above the alloy’s transformation temperature. Thus, the eyeglass frames of claim 5 exhibit the ability to resist permanent deformation to the extent of “3% heat-recoverable shape-memory” and “at least 3% elasticity” over a useful ambient temperature range.
II.
CVI and Marchon brought their action against Tura following reexamination of the ’955 patent. In due course, after Marcolin and Rothandberg were joined as plaintiffs, with the consent of the parties the case was tried before a jury with a United States magistrate presiding. At trial, CVUMarchon asserted that nine Tura frame models infringed claims 1-3, 5, and 6 of the ’112 and ’955 patents. At the close of the evidence, Tura moved for JMOL pursuant to Fed.R.Civ.P. 50(a), based upon its construction of the elasticity limitations of the claims. Tura argued that the Turaflex frames did not meet those limitations, as properly construed. Tura’s motion was denied.
Tura also asked the magistrate to instruct the jury as follows with respect to the meaning of the elasticity limitation in claim 1 of the ’955 patent:
The Defendants contend that “greater than 3% elasticity” means that when a force is applied to the frame to stretch it to a length at least 3% greater than its original length, and then the force is taken away, the frame will return to its original length. Moreover, the frame must exhibit this complete recovery from being stretched 3% ... whether the stretch and release is at any temperature from -20 degrees C to +40 degrees C.
*1151Tura sought a similar instruction with respect to the 4% elasticity limitation of claim 1 of the ’112 patent.
As far as claim 5 of the ’955 patent was concerned, Tura asked the magistrate to give the jury the following instruction:
[T]he Defendants contend that the phrases “3% heat recoverable shape-memory” and “3% elasticity” means [sic.] that when a component is stretched at room temperature to a length 3% greater than its original length, it will return to its original length when the stretching force is removed. However, if the component is overstretched at room temperature so that upon removal of the stretching force, the frame component remains 3% longer than when it started, then applying heat such as hot water, will cause the frame component to return to its original length.
Tura sought a similar instruction with respect to claim 5 of the T12 patent.
The magistrate declined to instruct the jury on the meaning of the claims. Rather, he only gave the jury general instructions relating to claim construction. The magistrate informed the jury that it should look to the ordinary meanings of the words used in the claims and that it should consider the patent specifications and the prosecution histories. The magistrate also instructed the jury that it should construe the claims as they would be construed by one of ordinary skill in the art, unless the specifications indicated that the inventors intended other meanings. After finding Tura LP liable for infringement of the asserted claims and Bro-dart Co. liable for inducing infringement, the jury awarded the damages noted above. CVI/Beta Ventures, 905 F.Supp. at 1204.
Following the jury verdict, Tura renewed its JMOL motion pursuant to Fed.R.Civ.P. 50(b). Id. at 1175. Tura argued that the Tura frames did not infringe the two patents because they did not meet the elasticity limitations of the claims at issue. Tura based its argument on the construction of those limitations that it had advanced in its proposed jury instructions. The magistrate described as follows CVl/Marchon’s and Tura’s positions on the claim construction issue:
The difference in the definitions of elasticity advanced by the parties concerns the behavior exhibited when the stresses on the rods are released. Plaintiffs assert that the term 3% elasticity refers to the ability of a metal item to “spring back” or recover in an amount equal to 3% of its original length after being subjected to stress. For example, to exhibit 3% elasticity as interpreted by plaintiffs, a metal rod 100 centimeters in length, after being stretched to a length of 105 centimeters when stressed, must, when the stress is released, “spring back” by an amount equal to 3% of its original length, or 3 centimeters, to a length of 102 centimeters.
According to defendants, 3% elasticity requires that a metal item return precisely to its original shape after being strained by at least 3% of its original length. In other words, to exhibit 3% elasticity as interpreted by defendants, a rod 100 centimeters in length, after being stretched to a length of 103 centimeters when stressed, would be required to recover fully and return spontaneously to its original length of 100 centimeters when the stress was released. (The behavior of the rod upon the release of stress after having been pulled to 105 centimeters is irrelevant to defendants’ interpretation.) In contrast, while plaintiffs’ definition would encompass full recovery from a 3% strain as one example of 3% elasticity, plaintiffs contend that any “spring back” of . at least 3%, whether or not the component returns to its original shape, is sufficient to meet the element of the patent claims.
Id. at 1177.
Under Tura’s construction of the term “3% elasticity” what is required is that an eyeglass frame component completely recover from a strain of 3%. Thus, according to Tura, the percentage appearing before the word “elasticity” refers to the amount of strain (or stretch) to which the component is subjected. Under Tura’s construction, as long as a 100 centimeter rod can fully return to its original shape when stretched to 103 centimeters, it is of no consequence whether the same rod returns by 2, 3, or 5 centimeters toward its original shape when pulled to *1152105 centimeters. Therefore, under Tura’s claim construction, “greater than 3% elasticity” in claim 1 of the ’955 patent means that a 100 centimeter rod must fully return to its original shape when stretched to, say, 103.5 centimeters. Under CVI/Marchon’s claim construction, “greater than 3% elasticity” means that a 100 centimeter rod stretched 3.5% to 103.5 centimeters must recover from that 3.5 centimeter stretch to the extent of at least some increment greater than 3 centimeters, for example, 3.1 centimeters. It need not, however, fully recover to 100 centimeters in length, as it must under Tura’s construction.
The magistrate rejected Tura’s claim construction argument. After examining the language of the claims, the specifications, the prosecution history, and extrinsic evidence in the form of expert testimony, he concluded that the term “3% elasticity is properly construed to refer to the behavior exhibited by a component which, when strained by a deforming force to a length in excess of 3% of its original length, springs back by at least 3% of its original length when the force is released.” Id. at 1182-83. Noting that Tura’s JMOL motion was premised on the claim construction which he had rejected, the magistrate denied the motion. Id.
DISCUSSION
I.
Judgment as a matter of law against a party is appropriate when “a party has been fully heard on an issue and there is no legally sufficient evidentiary basis for a reasonable jury to find for that party on that issue____” Fed.R.Civ.P. 50(a)(1). We review the district court’s decision to deny Tura’s JMOL motion de novo by reapplying the JMOL standard. Maxwell v. J. Baker, Inc., 86 F.3d 1098, 1104, 39 USPQ2d 1001, 1004 (Fed.Cir.1996); Markman v. Westview Instruments, Inc., 52 F.3d 967, 975, 34 USPQ2d 1321, 1326 (Fed.Cir.1995) (in banc), aff'd, — U.S. -, 116 S.Ct. 1384, 134 L.Ed.2d 577 (1996).
Determining whether a patent claim has been infringed requires a two-step analysis: “First, the claim must be properly construed to determine its scope and meaning. Second, the claim as properly construed must be compared to the accused device or process.” Carroll Touch, Inc. v. Electro Mechanical Sys., Inc., 15 F.3d 1573, 1576, 27 USPQ2d 1836, 1839 (Fed.Cir.1993). Claim construction is a question of law which we review de novo. Markman, 52 F.3d at 988, 34 USPQ2d at 1337. The application of a properly construed claim to an accused device is a question of fact which, following a jury trial, we review for substantial evidence. Genentech, Inc. v. Wellcome Found., Ltd., 29 F.3d 1555, 1565, 31 USPQ2d 1161, 1168-69 (Fed.Cir.1994).
II.
As seen above, the claim construction issue in this case centers on the “elasticity” limitations of independent claims 1 and 5 of the ’112 and ’955 patents. On appeal, Tura advances the same argument it raised before the district court. Thus, it contends that the “greater than 3% elasticity” limitation in claim 1 of the ’955 patent describes the ability of an eyeglass frame component to return completely and spontaneously to its original shape after being stretched by more than 3% of its original length. Under Tura’s construction of claim 1, the concept of “elasticity” requires complete recovery with zero residual strain — at least up to the point at which the component is strained to such an extent that it exceeds what we may call its “elastic limit.” When the component is stretched to such an extent that it exceeds its elastic limit, it is permanently deformed. CVI/Marchon urges that the district court’s claim construction — under which complete recovery to zero is not required — is correct.
In determining the proper construction of a claim, “the court should look first to the intrinsic evidence of record, ie., the patent itself, including the claims, the specification, and if in evidence, the prosecution history.” Vitronics Corp. v. Conceptronic, Inc., 90 F.3d 1576, 1582, 39 USPQ2d 1573, 1576 (Fed.Cir.1996). We have described such intrinsic evidence as “the most significant source of the legally operative meaning of claim language.” Id. Extrinsic evidence *1153(e.g., expert testimony) also may be used in claim construction. Id. However, when “an analysis of the intrinsic evidence alone will resolve any ambiguity in a disputed claim term,” it is improper to rely on extrinsic evidence. Id. at 1583, 39 USPQ2d at 1577.
As used in the ’112 and the ’955 patents, respectively, 4% and 3% “elasticity” are technical terms. We have stated that “[a] technical term used in a patent document is interpreted as having the meaning that it would be given by persons experienced in the field of the invention, unless it is apparent from the patent and the prosecution history that the inventor used the term with a different meaning.” Hoechst Celanese Corp. v. BP Chems. Ltd., 78 F.3d 1575, 1578, 38 USPQ2d 1126, 1129 (Fed.Cir.1996). Thus, “it is always necessary to review the specification to determine whether the inventor has used any terms in a manner inconsistent with their ordinary meaning.” Vitronics, 90 F.3d at 1582, 39 USPQ2d at 1577.
The claims of the ’112 and ’955 patents do not define the terms 4% and 3% “elasticity.” We thus turn to the specifications.5 In that regard, the patent drawings are highly relevant in construing the elasticity limitations of the claims. Of particular importance are Figures 2F, 2G, and 2H, which are reproduced below:
*1154[[Image here]]
The figures are graphs depicting, in the form of stress-strain curves, stress-strain behavior at different temperatures (T). In each graph, the stress (or the pulling force) to which the alloy being tested is subjected is measured on the vertical axis. The horizontal axis plots the strain which the alloy being tested exhibits at any given amount of stress. In other words, the horizontal axis shows the degree to which the alloy stretches when it is pulled and the degree to which it returns to its original shape when the pulling force is reduced or released. ’112, col. 3, lines 20-27; ’955, col. 3, lines 43-54. Figures 2F and 2H show, up to a certain limit, complete release of stress causing strain to return to zero. Although for different temperature ranges, Figures 2F and 2H both show an alloy (i) being subjected to stress resulting in a strain of approximately 6% and (ii) then returning to its original shape when the stress is released. Figure 2G shows partial return of strain to zero. Figure 2G shows an alloy (i) being subjected to stress resulting in a strain of between 6 and 8% and (ii) then returning to between 3 and 4% of its original shape when the stress is released. Figure 2G also shows strain returning to zero upon the application of heat. Referring to Figure 2H above, the specifications for both the ’112 and the ’955 patent state as follows: “Thus, throughout the temperature region of inter*1155est for eyeglass frames, the component acts completely elastically up to strains of 6% or more.” ’112, col. 5, lines 24L-26; ’955, col. 5, line 68-eol. 6, line 3. The specification for the ’955 patent continues in the immediately following paragraph to give an example of an embodiment of claim 1. The example states that a “suitable component” will “support a stress of over 150 ksi ... at 4% tensile strain and show complete elastic spring back at room temperature.” ’955, col. 6, lines 5-14. The specification concludes the discussion of the indicated example by stating that “[f]rom the above, it can be seen that an eyeglass frame may be constructed with portions thereof exhibiting optimized elasticity, such portions ... having greater than 6% elasticity over a temperature range from -20°C. to +40°C.” ’955, col. 6, lines 19-25.
The specification for the ’112 patent contains no examples of the claim 5 embodiment. Each example of a claim 5 component in the ’955 patent specification exhibits some degree of retained strain, which is removed by heating. The first example is of a component which, after having been subjected to a 7.5% strain and then having the strain released, “gives 3.7% elastic spring back and imparts 3.8% shape-memory recovery when heated.” ’955, col. 6, lines 60-62. The second example is of a component which, when subjected to a 7.25% strain and then having the strain released, “gives 5.0% elastic spring back and imparts 2.25% shape-memory recovery” when heated. ’955, col. 7, lines 3-5. Figure 2G above depicts the behavior of an alloy possessing both “elasticity” and “heat-recoverable shape-memory.”
We turn next to the prosecution history. We have stated that “arguments made during prosecution regarding the meaning of a claim term are relevant to the interpretation of that term in every claim of the patent absent a clear indication to the contrary.” Southwall Techs. Inc. v. Cardinal IG Co., 54 F.3d 1570, 1579, 34 USPQ2d 1673, 1679 (Fed. Cir.), cert. denied, — U.S. -, 116 S.Ct. 515, 133 L.Ed.2d 424 (1995). Prosecution history “limits the interpretation of claim terms so as to exclude any interpretation that was disclaimed during prosecution.” Id. at 1576, 34 USPQ2d at 1676. See also Biodex Corp. v. Loredan Biomedical, Inc., 946 F.2d 850, 863, 20 USPQ2d 1252, 1262-63 (Fed.Cir.1991) (a particular interpretation of a claim can be disclaimed by the inventor during prosecution).
We consider first statements made by the inventors during the prosecution of the ’112 patent. Claims 1-16 in the application for the ’112 patent originally were rejected by the examiner as being unpatentable over various Japanese patent references. In that regard, what was then an amended version of claim 1 was rejected under 35 U.S.C. § 103 as obvious over Japanese reference P5689717 (the ’717 reference). The ’717 reference teaches eyeglass frames having at least a portion made from shape-memory alloy. It also teaches work-hardening as part of the fabrication process. The applicants amended claim 1 again, thereby putting it into its present form. Thereafter, in the Remarks portion of the Response to Office Action, the applicants sought to distinguish their further-amended claim 1 over the ’717 reference. The applicants argued that claim 1 was “specific to Applicants’ first embodiment of frame components with optimized elasticity.” Claim 1, they continued, “recites that the component will have at least 4% elasticity over the expected usage temperature range of -20 °C. to +40 °C. and will have a low effective elastic modulus which gives it a soft, springy feel.” Applicants stated that “[t]he figure which illustrates this embodiment is Fig. 2H.” Figure 2H, of course, depicts full return to zero after the release of stress.
In 1990, both CVI and Tura LP sought reexamination of claim 1 of the ’955 patent. The reexamination focused on an article published in 1982 in Journal de Physique, a French publication. The article was by O. Mercier and E. Torok and was titled “Mechanical Properties of the Cold-Worked Martensitic NiTi Type Alloys” (“Mercier”). Among other things, Mercier discusses the pseudoelastic properties of nickel-titanium alloys. Tura LP argued that claims 1-4 of the ’955 patent would have been obvious over Mercier because “Mercier et al. fully disclose the nickel-titanium based shape-memory alloy defined in claims 1-4.”
*1156In making its obviousness argument, Tura LP pointed to Figure 2H from the ’955 patent. It asserted that Figure 2H illustrated “[t]he behavior of the alloy in the work-hardened pseudoelastie state” and that “[i]n this condition, the alloy is claimed to have greater than 3% elasticity over a temperature range from -20 °C. to +40 °C.” It then turned to Figure 4 of Mercier, which is shown below:
[[Image here]]
[App. 7031]
Tura LP argued that curve EC’DE of Mercier “perfectly anticipates Fig. 2H of the Zider et al. patent.” That curve, it elaborated, “shows an elasticity, between points C’ and E of 8.1%, which anticipates ‘greater than 3% elasticity’ set forth in the patent.” Thus, in the reexamination proceeding, Tura advanced the construction of the term “greater than 3% elasticity” which the district court adopted and which CVI/Marchon urges. This is so because Figure 4 of Mercier clearly shows (i) a deforming stress resulting in a strain of 11%, (ii) release of that stress, and (iii) a recovery of 8.1% (which is a recovery of more than 3% but not a return to zero.)
CVT responded by seeking to distinguish claim 1 of the ’955 patent over Mercier. Noting that Figure 4 of Mercier exhibited a “3% retained strain” when the deforming stress was removed, CVI argued that
Figure 2H of the ’955 patent shows how the work-hardened pseudoelastic material *1157covered by claim 1 will behave at temperatures between Ms and Md and it can clearly be seen that the material is totally elastic when a stress of about 115 ksi is applied and removed. The sample of Mercier shown in Figure 4 clearly does not exhibit the properties shown in Figure 2H of the ’955 patent when a load of 1150 N/mm2 is applied and removed.
CVI also argued that a person of ordinary skill in the art would not consider the test at 100 °C. shown in Figure 4 of Mercier as being pertinent in designing eyeglass frames to be worn at temperatures of -20 °C. to +40 °C. Continuing in this vein, CVI asserted that Mercier provided no information as to what the properties of the sample tested at 100 °C. in Figure 4 would be in the temperature range of -20 °C. to +40 °C. Under these circumstances, CVI stated, “Mercier does not provide any suggestion of the unique combination of properties which can be obtained with nickel-titanium alloys in the work-hardened pseudoelastic state.” CVI then contrasted the meaning of “8.1% elasticity” in Mercier with the “3% elasticity” of claim 1:
[T]he third party [Tura] contends that the 8.1% elasticity shown in Figure 4 anticipates the 3% elasticity of claim 1. The 8.1% elasticity shown in Figure 4 of Mercier eomes at the expense of a 3% retained strain, that is, the material will not return to its original shape. The material covered by Claim 1 of the ’955 patent exhibits at least 3% elasticity over the temperature range -20 to + 40°C. The elasticity set forth in Claim 1 is attainable prior to the onset of permanent deformation. Otherwise, the recited elasticity properly [sic] would be meaningless. That is, elasticity measurements are used to indicate how much a material can be elastically deformed. Once a material is plastically deformed, it no longer will return to its original shape. With respect to the data shown in Figure 4 of Mercier, if the sample prior to testing corresponded to a finished part, after applying a load of 1150 N/mm2 and releasing the load, the part would be deformed such that it exhibited 3% retained strain. In the context of eyeglass frames, a frame which is deformed by 3% strain yet exhibits 8.1% elasticity is of no value since the 3% permanent deformation would result in significant distortions in the original shape of the frame.
Finally, in its discussion of Mercier, CVI referred to excerpts from Mechanical Metallurgy by G.E. Dieter, Jr. (McGraw-Hill Book Co., 1961), as “explaining ‘elastic behavior’, ‘plastic deformation’____” The Dieter reference states in part:
Experience shows that all solid materials can be deformed when subjected to external load. It is further found that up to certain limiting loads a solid will recover its original dimensions when the load is removed. The recovery of the original dimensions of a deformed body when the load is removed is known as elastic behavior. The limiting load beyond which the material no longer behaves elastically is the elastic limit. If the elastic limit is exceeded, the body will experience a permanent set or deformation when the load is removed. A body which is permanently deformed is said to have undergone plastic deformation.
In affirming the patentability of the claims of the ’955 patent, the examiner rejected Tura LP’s argument that Mercier anticipated claim 1. Considering Mercier, the examiner noted that Mercier involved a different alloy than the one- in the ’955 patent and that it exhibited “desired properties in a temperature range outside that limited by the claim.” Under these circumstances, the examiner wrote, “the materials of Zider et al are not anticipated by Mercier et al. Further, since the results of a change in materials are unpredictable, there is no suggestion that such a change would have been obvious, or that such a change would produce desirable results.”
III.
Having reviewed the patent specifications and the prosecution histories,6 we con-*1158elude that the district court erred in its claim construction. We conclude that, as used in claims 1 and 5 of the ’112 and the ’955 patent, the word “elasticity” refers to the ability of an eyeglass frame component to return completely and spontaneously to its original shape after it is subjected to stress and the stress is removed, provided that the stress to which the component is subjected does not cause the component to exceed its elastic limit. We also conclude that the percentages appearing before the word “elasticity” in claims 1 and 5 of the patents refer to the amount of strain to which the eyeglass frame component is subjected. Thus, “greater than 3% elasticity” in claim 1 of the ’955 patent means that a 100 centimeter rod must fully return to its original shape when stretched to some length over 103 centimeters. By the same token, “at least 3% elasticity” in claim 5 of the ’955 patent means that the same rod must fully return to its original shape when stretched to any length under 103 centimeters.
Even standing alone, the specifications suggest this construction. Figures 2F and 2H above plainly show complete return to zero after the release of stress. At the same time, the specifications for both patents refer to Figure 2H and use the term “completely elastically” when referring to the desired behavior of eyeglass frame components. ’112, col. 5, lines 24-26; ’955, col. 5, line 68-col. 6, line 3. Likewise, the specification for the ’955 patent states that a “suitable component” will show “complete elastic spring back at room temperature.” ’955, col. 6, lines 5-14. Finally, the reference to complete elasticity “up to strains of 6% or more” in both patents, and the reference to a similar degree of elasticity “at 4% tensile strain” in the ’955 patent support the conclusion that the percentages appearing before the word “elasticity” in the patents indicate the amount of strain to which the eyeglass frame component is subjected.
CVI/Marchon correctly notes that the specification for the T12 patent recites that Figures 2F and 2H “represent the ideal characteristics for use in eyeglass components,” and that the specification for the ’955 patent states that Figures 2F and 2H represent “the ideal characteristics for use in purely elastic eyeglass components ....” ’955, col. 5, lines 19-21. CVI/Marchon argues that “neither the ’112 nor the ’955 patent claims are limited to the ‘optimum’ embodiment.”
CWMarchon is correct that, as a general matter, the claims of a patent are not limited by preferred embodiments. See Sjolund v. Musland, 847 F.2d 1573, 1581, 6 USPQ2d 2020, 2027 (Fed.Cir.1988) (noting that while claims are to be interpreted in light of the specification, it does not follow that limitations from the specification may be read into the claims); Constant v. Advanced Micro-Devices Inc., 848 F.2d 1560, 1571, 7 USPQ2d 1057, 1064 (Fed.Cir.1988) (“Although the specification may aid the court in interpreting the meaning of disputed language in the claims, particular embodiments and examples appearing in the specification will not generally be read into the claims.”). However, as discussed above, through statements made during prosecution or reexamination an applicant for a patent or a patent owner, as the case may be, may commit to a particular meaning for a patent term, which meaning is then binding in litigation. That, we believe, is what happened here. Statements made during prosecution of the ’112 patent and during reexamination of the ’955 patent compel the conclusion that it was the intention of the inventors that, as used in claims 1 and 5, “elasticity” means total recovery prior to the onset of permanent deformation.
We note first the statement made during the prosecution of the ’112 patent that amended claim 1 of the application was “specific to Applicants’ first embodiment of frame components with optimized elasticity” and that “[t]he figure which illustrates this embodiment is Fig. 2H.” Figure 2H clearly shows return to zero after the release of deforming stress. Thus, the inventors clearly tied “elasticity” in claim 1 of the ’112 patent to the concept of total recovery. The statements in the ’112 and ’955 specifications that Figures 2F and 2H represent “ideal characteristics” are not inconsistent with this meaning for “elasticity.”
*1159We find compelling the statements made by CVI during reexamination of the ’955 patent to overcome Tura LP’s arguments that Figure 4 of Mercier rendered claim 1 obvious. As seen above, CVI responded that “Figure 2H of the ’955 patent shows how the work-hardened pseudoelastic material covered by claim 1 will behave at temperatures between Ms and Md and it can clearly be seen that the material is totally elastic when a stress of about 115 ksi is applied and removed.” CVI continued that “[t]he sample of Mercier shown in Figure 4 clearly does not exhibit the properties shown in Figure 2H of the ’955 patent when a load of 1150 N/mm2 is applied and removed.” In our view, this statement could hardly be clearer. The only meaning we are able to draw from it is that CVI was telling the examiner, “Claim 1 relates to an eyeglass frame component where there is total spring back to zero after the removal of a deforming force, as seen in Figure 2H of the patent. That is not what is shown in Figure 4 of Mercier, because that figure shows 3% retained strain.” That also is the thrust of CVTs additional statement that “[t]he 8.1% elasticity shown in Figure 4 of Mercier comes at the expense of a 3% retained strain, that is, the material will not return to its original shape.” In addition, CVI emphasized to the examiner that an eyeglass frame which was “deformed by 3% strain” would be “of no value” even if, as depicted in Figure 4 of Mercier, it exhibited “8.1% elasticity.”
In support of its claim construction, CVI/Marehon points to claim 5. It argues that, as used in claim 5 in both the T12 and the ’955 patents, the term “at least 3% elasticity” cannot require “full recovery.” This is so, CVI/Marehon asserts, because claim 5 shows a combination of “3% heat-recoverable shape-memory” and “3% elasticity.” Put another way, if a component is stretched and then does not recover through its “elasticity” alone, unaided by “heat-recoverable shape-memory,” then necessarily “elasticity” cannot mean full recovery, as Tura urges. According to CVI/Marehon, claim 5 and the related portions of the ’955 specification thus provide support for its argument with respect to the construction of claim 1 of the patents.
This argument is flawed. First, CVI/Marchon’s proffered meaning of “elasticity” is squarely at odds with the meaning of the term that emerges from our examination of the specifications and the prosecution and reexamination histories, as detailed above. CVI/Marehon cannot now turn its back on the meaning of “elasticity” which it embraced for claim 1 during the prosecution of the T12 patent and during the reexamination proceeding relating to the ’955 patent. At the same time, we are obliged to construe the term “elasticity” consistently throughout the claims. See Fonar Corp. v. Johnson & Johnson, 821 F.2d 627, 632, 3 USPQ2d 1109, 1113 (Fed.Cir.1987) (the meaning of a term in a claim must be defined in an manner that is consistent with its appearance in other claims in the same patent).
Moreover, contrary to CVl/Marchon’s suggestion, it is not inconsistent to have “elasticity,” as it appears in claim 5 of both patents (“at least 3% elasticity”), mean total recovery. The eyeglass frame of claim 5 is different from the frame of claim 1. Taking the ’955 patent, the frame of claim 1 is in “the work-hardened pseudoelastie metallurgical state.” It must show complete recovery from “greater than” 3% strain “over a temperature range from -20 °C. to + 40 °C.” The elastic properties of a claim 1 frame are depicted in Figure 2H of the ’955 specification. In Figure 2H, the temperature range limitation is indicated by the symbol Ms < T2 < Md, where Ms and Md represent the lower and upper ends of the useful ambient temperature range, respectively. According to the inventors, claim 1 frames embody the concept of “optimized elasticity.”
The frame of claim 5, on the other hand, is in “the work-hardened and heat-treated condition.” It must show “a minimum of 3% heat-recoverable shape-memory, a yield strength greater than 30,000 psi and at least 3% elasticity.” The elastic properties of a claim 5 frame are depicted in Figure 2G of the ’955 specification. Figure 2G shows “a combination of elastic and shape-memory properties” at the low end of the useful ambient temperature range (T < Ms). ’112, col. 3, lines 23-24; ’955, col. 3, lines 47-50. During the prosecution of the ’112 patent, the *1160inventors stated that “[t]he memory plus elasticity embodiment [of the invention] is shown in Figure 2G.” The inventors continued by describing the advantage which flows from having heat-recoverable shape-memory in addition to work-hardening: “Conversely, if the same material were left in a fully work-hardened condition, the properties shown in Fig. 2F would result, and the component would be very springy but would not give a complete shape-memory recovery if deformed past the component’s elastic limit.” (Figure 2F shows stress at T < Ms.) In short, the heat-recoverable shape-memory limitation of claim 5 speaks to the situation in which an alloy is stressed to the point that its inherent elasticity is not enough to enable it to return to its original shape when the deforming force is removed. In other words, the alloy’s elastic limit is exceeded. As noted above, this retained strain is seen in the examples in the ’955 specification which discuss the embodiment of claim 5. That is not the case in claim 1. As CVI stated during reexamination of the ’955 patent, “[t]he elasticity set forth in Claim 1 is attainable prior to the onset of permanent deformation.”
A claim 5 component has two distinct attributes at T < Ms, elasticity and heat-recoverable shape-memory. Thus, the component can fully recover its original shape if it is subjected to a deforming force which stretches it up to 3% of its original length. Up to that point, only the component’s elasticity properties are in evidence. If, however, the component is subjected to a deforming force which stretches it more than 3% of its original length, so that its elastic limit is exceeded, as in the examples in the ’955 specification, it draws on its shape-memory properties when heat is applied, thereby enabling it to return to its original shape. The meaning given to “elasticity” by the inventors works in both claims 1 and 5.
Finally, our interpretation of the claims is consistent with and furthers the purpose of the invention. Eyeglass frames normally are carefully fitted to the wearer’s particular facial configuration. If the fit is not correct, the eyeglasses are unlikely to perform their vision-correcting function properly or to be comfortable or acceptable to the wearer. When the shape of the frames is suddenly altered, such as by dropping them or sitting on them, distortion of the shape makes the frames less useful or attractive to the wearer. A frame whose elasticity enables it to correct the entire change by returning the frame to its original shape would be the most useful and most desired by the wearer. In these circumstances, it is reasonable to conclude that the references to 3% and 4% “elasticity” in the claims was intended to describe an eyeglass frame that would have that desirable and important property. In construing claims, the problem the inventor was attempting to solve, as discerned from the specification and the prosecution history, is a relevant consideration. See Applied Materials v. Advanced Semiconductor Materials, 98 F.3d 1563, 1573, 40 USPQ2d 1481, 1488 (Fed.Cir.1996).7
IV.
What we must decide now is whether, in view of the correct construction of the claims, we should reverse the judgment of infringement in favor of CVI/Marehon or should *1161vacate the judgment and remand the case for a new trial on the infringement issue. For the reasons which follow, we conclude that, because of CVI/Marchon’s failure of proof under the correct claim construction, Tura is entitled to a judgment of noninfringement as a matter of law.
The question before us is whether any jury could reasonably have found that the accused Tura frames infringed the claims of the ’112 and ’955 patents as properly construed, either literally or under the doctrine of equivalents. Exxon Chem. Patents, Inc. v. Lubrizol Corp., 64 F.3d 1553, 1558, 35 USPQ2d 1801, 1805 (Fed.Cir.1995), cert. denied, — U.S. -, 116 S.Ct. 2554, 135 L.Ed.2d 1073 (1996). In that regard, the Supreme Court has stated that “[I]f the evidence presented in the first trial would not suffice, as a matter of law, to support a jury verdict under the properly formulated defense, judgment could properly be entered for the respondent at once, without a new trial.” Boyle v. United Techs. Corp., 487 U.S. 500, 513, 108 S.Ct. 2510, 2519, 101 L.Ed.2d 442 (1988). The question of whether the evidence of record is sufficient to create an issue of fact for the jury is itself a question of law, which we decide. Exxon, 64 F.3d at 1559, 35 USPQ2d at 1805.
In order for there to be infringement, each and every limitation set forth in a patent claim must be found in the accused product, either literally or under the doctrine of equivalents. Lemelson v. United States, 752 F.2d 1538, 1551, 224 USPQ 526, 533 (Fed.Cir.1985). The patentee bears the burden of proving infringement by a preponderance of the evidence. Laitram Corp. v. Rexnord, Inc., 939 F.2d 1533, 1535, 19 USPQ2d 1367, 1369 (Fed.Cir.1991). For our purposes, we are concerned solely with the elasticity limitations of the claims. If those limitations are not met in the accused frames, CVI/Marchon’s allegations of infringement fail.
As noted above, CVI/Marchon alleged that certain Tura frames infringed claims 1-3, 5 and 6 of both the ’112 and the ’955 patents. For the most part, CVI/Marchon limited its proof at trial to literal infringement. It did, however, assert infringement under the doctrine of equivalents with respect to the “greater than 4% elasticity” limitation in claims 1-3 of the ’112 patent and with respect to the heat-treatment limitation in claim 5 of the patents.
On the record before us, there is no evidence that the accused Turaflex frames meet the elasticity limitations of claims 1 and 5 of the patents, as those limitations are properly construed. Turning first to the “greater than 3% elasticity” limitation of claim 1 of the ’955 patent, we note that none of the accused frames, when tested, showed full recovery to its original shape after the release of stress. Indeed, Dr. Daniel Beshers, CVI/Marchon’s expert witness, stated that, assuming 3% elasticity meant full recovery, “many of [the Tura frames] are not in that exact 3 percent recovery.” He further testified that when the frames were tested at -20 °C., the low end of the claimed temperature range, none of them showed full recovery. CVI/Marchon also failed to come forward with evidence demonstrating that the Tura frames exhibit “at least 3% elasticity” as recited in claim 5 of the ’112 and the ’955 patents as that limitation is properly construed. Thus, on the record before us, no reasonable jury could find that the Tura frames literally infringe claims 1 and 5 of the patents.
In order to establish infringement under the doctrine of equivalents, CVI/Marchon must show evidence in the record to support the contention that the Tura frames exhibit the equivalent of “at least 3% elasticity” as recited in claim 5 of the patents, or “greater than 3%” or “4% elasticity” as recited in claim 1 of the ’955 and the ’112 patents, respectively. CVI/Marchon has failed to point to any testimony or evidence in the record regarding the technical issue of equivalency according to the proper claim construction. See Lemelson v. General Mills, Inc., 968 F.2d 1202, 1208, 23 USPQ2d 1284, 1289 (Fed.Cir.1992) (patentee’s failure at trial to identify elements in the accused device that correspond to two limitations in the claim indicates that no reasonable jury could find infringement, either literally or by equivalents). Therefore, no reasonable jury *1162could find infringement under the doctrine of equivalents.
In the district court, CVI/Marchon was on notice that adoption of Tura’s claim construction was a distinct possibility. Prior to trial, in ruling on motions for summary judgment, the court agreed with Tura that “3% elasticity” meant complete recovery to the frame’s original length when the frame was subjected to 3% strain. Nevertheless, CVI/Marchon confined its proof at trial to evidence purporting to establish infringement under its construction of the claims. That construction was not accepted by the court, however, until the magistrate denied Tura’s JMOL motion at the close of the evidence and declined to charge the jury in accordance with Tura’s profferred instructions. Significantly, on appeal, CVI/Marchon has not advanced an alternative argument to the effect that, even under the claim construction advanced by Tura, there is substantial evidence in support of the jury’s verdict of infringement.
Under the circumstances of this case, the proper course is reversal of the judgment of infringement, rather than a vacatur of the judgment and a remand for a new trial. That is the teaching of Exxon. In that case, Exxon Chemical Patents, Inc. had sued Lubrizol Corporation for literal infringement of Exxon’s patent relating to a lubricating oil composition. At trial, Exxon and Lubrizol advanced competing claim constructions. Exxon sought to prove infringement under Lubrizol’s claim construction, not knowing which claim construction the trial court would adopt. The court charged the jury in accordance with Exxon’s construction. The jury returned a verdict of infringement and Lubrizol appealed. On appeal, we determined that the correct claim construction was “but a slight variance” from that urged by Lubrizol. Exxon, 64 F.3d at 1560, 35 USPQ2d at 1807. Since Exxon never identified the proof that the accused product infringed under Lubrizol’s claim construction, we concluded that remand for a second trial was not necessary. Instead, we reversed because we concluded that no jury could reasonably find that Lubrizol’s accused product literally infringed the claims of the patent as properly construed. We stated: “Exxon’s error was in failure of proof as to the claimed amounts [of one of the ingredients of the lubricating oil that was the subject of the claims], without which it could not prove infringement under Lubrizol’s claim meaning.” Id. at 1561, 35 USPQ2d at 1807.
This appeal presents an even more compelling case for reversal than Exxon. Here, the correct claim construction is the same as the one advanced by Tura, and CVI/Marchon has not pointed to any evidence in the record as supporting the judgment of inftingement when the claims are correctly construed. Finally, on appeal, CVI/Marchon does not ask for a new trial in the event that we agree with the claim construction urged by Tura. Under these circumstances, it is clear that Tura is entitled to a judgment of noninfringement as a matter of law.
CONCLUSION
The elasticity limitations of independent claims 1 and 5 of the ’112 and the ’955 patents require full recovery after the release of stress. At the same time, there is no evidence of record based upon which a reasonable jury could find that, as correctly construed, the asserted claims (1-3, 5, and 6 of both patents) are met in the accused Tura frames, either literally or under the doctrine of equivalents. Thus, there can be no infringement. Accordingly, the judgment of infringement in favor of CVI/Marchon is reversed and the award of damages and the injunction entered against Tura are vacated. The ease is remanded to the district court for entry of judgment of noninfringement in favor of Tura and for further proceedings in connection with Tura’s pending counterclaims.
COSTS
Each party shall bear its own costs.
REVERSED and REMANDED.
. Suit also was brought against Arthur Brody and Bracken Opticians for inducing infringement. The district court dismissed the claim against Brody on venue grounds, and no judgment was entered against Bracken. Neither party is involved in this appeal.
. No validity issues are raised on appeal.
. —20 °C. is equal to -4 0 F. +40 »C is equal to + 104 »F.
. The term “transformation temperature” refers to the temperature region at which an alloy exhibits a change in its elastic properties, resulting from a change in the crystalline structure of the alloy at the microscopic level.
. The specification for the '112 patent contains a typographical error. Under the heading “DESCRIPTION OF THE PREFERRED EMBODIMENT," the reference in column 4 at line 64 to “FIG. 2g" should be to "FIG. 2f." This is clear from the context.
. We view this as a case in which reliance on extrinsic evidence (e.g., expert testimony) is not necessary.
. In CVI/Beta Ventures, Inc. v. Custom Optical Frames, Inc., Nos. 96-1070, 95-1486, 1996 WL 338388 (Fed.Cir. June 19, 1996), a different panel of this court, in a nonprecedential opinion, upheld the grant of a preliminary injunction against infringement of the '112 and '955 patents. See CVI/Beta Ventures, Inc. v. Custom Optical Frames, Inc., 893 F.Supp. 508 (D.Md.1995). In so doing, the panel affirmed an interpretation of the 3% elasticity limitation which did not restrict the claim to complete recovery. In its opinion, the panel stated that it could not conclude "in the context of the preliminary injunction proceedings that the district court erred in rejecting Custom Optical's proffered claim construction.” The panel noted, as well, that the parties would have the opportunity at the merits stage to expand their arguments and to present any additional arguments. In this appeal, we review a different trial court’s final claim construction as part of our review of the judgment on infringement. Therefore, unlike the earlier appeal, this appeal required us to construe the asserted claims based upon the final and complete record in the case. See International Communication Materials, Inc. v. Ricoh Co., Ltd., 108 F.3d 316, 319 (Fed.Cir.1997) ("We review only the district court’s tentative claim construction and its resulting finding on the likelihood of successfully proving infringement, which form an adequate basis for our affirming its denial of the preliminary injunction.”).