Opinion concurring in part and dissenting in part filed by Circuit Judge SCHALL.
MOORE, Circuit Judge.Perfect Surgical Techniques, Inc. (“PST”) appeals from a decision of the Patent Trial and Appeal Board (“Board”) invalidating claims 1, 4-6, 8, 9, 11, 12, 38, 41-44, 46, 47, and 49 of U.S. Patent No. 6,030,384 (“’384 patent”). For the reasons set forth below, we vacate and remand.
Background
This appeal arises from the Board’s inter partes review (“IPR”) decision determining, inter alia, that each of the claims at issue are anticipated or would have been obvious over Japanese Application Publication No. H10-33551 A (“JP ’661”). The Board determined that JP ’551 qualified as prior art to the ’384 patent under 35 U.S.C. § 102(a). First, the Board rejected PST’s argument that the Board could not rely on JP ’551 as filed because the petitioners, Olympus America, Inc. and Olympus Medical Systems Corp. (collectively, “Olympus”), failed to translate the bibliographic page containing the publication date. Second, the Board determined that PST failed to antedate JP ’551 because it had not proven that the inventor of the ’384 patent was reasonably diligent in reducing his invention to practice. After concluding that JP ’551 was prior art, the Board examined the prior art under its claim construction and determined JP ’551 anticipated or rendered obvious the claims at issue.
PST’s appeal addresses all three aspects of the Board’s decision. We have jurisdiction under 28 U.S.C. § 1295(a)(4)(A).
Discussion
A. Partial Translation of JP ’551
Pursuant to 37 C.F.R. § 42.63(b), the Board requires a party relying on a non-English document to submit “a translation of the document into English and an affidavit attesting to the accuracy of the translation.” PST argues that Olympus’ failure to translate the bibliographic page of JP ’551 rendered their submission defective and that the Board violated its own regulations by relying on Olympus’ partially-translated submission. PST argues that the bibliographic page contained the publication date, and since that page was not translated, Olympus did not present sufficient evidence for the Board to determine that JP ’551 was prior art to the ’384 patent.
The Board is not free to disregard its own regulations. Nor can it mandate compliance by only some parties. But in this case, even if the Board erred by accepting Olympus’ submission of JP ’551, its error is harmless.
*1007PST concedes that the only relevant information on the untranslated page is the publication date, and the date is discerna-ble from the face of the untranslated page. See J.A. 13 (citing Manual of Patent Examining Procedure § 901.05(a) (describing how to convert Japanese format publication date information to Gregorian calendar date information)). As noted by the Board, the translator’s certification further attested that JP ’551 was published' on February 10, 1998. J.A. 12-13. PST does not dispute that JP ’551 was published on this date, nor does it dispute the remainder of the translation. For these reasons, PST has not established that the Board’s acceptance of a not fully translated version of JP ’551 was anything other than harmless error.
B. Antedating JP ’551
1. Background
Pre-AIA section 102(g) allows a patent owner to antedate a reference by proving earlier conception and .reasonable diligence in reducing to practice.1 Monsanto Co. v. Mycogen Plant Sci, Inc., 261 F.3d 1356, 1362 (Fed. Cir. 2001). Reasonable diligence must be shown throughout the entire critical period, which begins just prior to the competing reference’s effective date and ends on the date of the invention’s reduction to practice. Id. at 1363.
Dr. Camran Nezhat, the inventor of the ’384 patent, hired James Heslin to draft and file the application that would issue as the ’384 patent. On January 28, 1998, Mr. Heslin sent an initial draft of the application to Dr. Nezhat. Mr. Heslin later filed the application with the Patent and Trademark Office (“PTO”) on May 1,1998. During this 93-day lapse of time—and only thirteen days after Dr. Nezhat received the initial draft application—JP ’551 published. To show diligence in reduction to practice, PST must prove that Dr. Ne-zhat was reasonably diligent during the critical period, which began on February 9, 1998, one day prior to JP ’551’s publication date, and ended on May 1, 1998, the date the invention was constructively reduced to practice by filing the application that resulted in the ’384 patent with the PTO.
Dr. Nezhat testified that he was reasonably diligent in reducing his invention to practice from February 9 to May 1, 1998. He testified that, although he did not recall all of his daily activities, he “diligently worked with Mr. Heslin between February I,1998 and May 1,1998, within the reasonable limits of my busy medical practice and teaching schedule[.]” J.A, 686. He testified that throughout these months, his medical practice required him to work roughly 80 hours per week, during which he would perform approximately four to six surgeries. Id. He testified that he submitted comments on the initial draft application on March 2, received questions from Mr. Hes-lin on or around March 4 and 12, participated in a conference with Mr. Heslin on March 16, and received a second draft application from Mr. Heslin on April 13. J.A. 686-87. He testified that he “again carefully reviewed the second draft” before it was finalized and filed on May 1. J.A. 687.
An inventor’s testimony regarding his reasonable diligence must be corroborated by evidence. Brown v. Barbacid, 436 F.3d 1376, 1380 (Fed. Cir. 2006). A “variety of activities” may corroborate an inventor’s testimony of reasonable diligence *1008and such corroborating evidence is considered “as a whole” under a rule of reason. Id. To corroborate Dr. Nezhat’s testimony, PST introduced a declaration from Mr. Heslin, letters between Dr. Nezhat and Mr. Heslin, and Mr. Heslin’s billing records.
Mr. Heslin recounted the same dates of activity as Dr. Nezhat and added that he reviewed and revised the initial draft application on April 7. J.A. 757-58. He characterized Dr. Nezhat as a “careful and serious client” who “communicated with me, in writing and in person, to finalize the application.” J.A. 758. He testified that he and Dr. Nezhat “put in a significant amount of effort and work into revising and finalizing the patent application between February 1 and May 1.” J.A. 758-59. He testified that during the relevant time period, he “maintained a very active practice” and typically worked 50 hours or more a week. J.A. 756, 758.
Mr. Heslin’s declaration appended a number of exhibits as support for his testimony. A March 12 letter from Mr. Heslin thanked Dr. Nezhat for his “facsimile of March 2, 1998, with comments on the initial draft.” J.A. 927. The letter asked for Dr. Nezhat’s “further comments on electrode surface area” and requested that they meet “to discuss finalizing the draft.” Id. Mr. Heslin’s billing records indicate he reviewed Dr. Nezhat’s comments on March 4 and that he met with Dr. Nezhat on March 16. J.A. 604. The billing records also indicate Mr. Heslin revised the draft application on April 7. Id. An April 13 letter from Mr. Heslin to Dr. Nezhat enclosed a second draft application and expressed that he “tried to reflect the comments you made when we last met.” J.A. 607. The letter asked for Dr. Nezhat’s “further comments so that [the application] can be finalized and filed in the near future.” Id. Mr. Heslin filed the application on May 1. J.A. 604, 634.
The Board accepted Dr. Nezhat’s testimony and corroborating evidence but determined that PST was not “sufficiently specific as to facts and dates” of Dr. Ne-zhat’s diligence during three portions of the critical period: (1) February 10 to March 1; (2) March 12 to March 15; and (3) April 13 to May 1. J.A. 21-22. The Board therefore concluded that PST failed to demonstrate Dr. Nezhat’s “continuous exercise of reasonable diligence for the entire critical period.” J.A. 22.
2. Discussion
Whether a patent antedates a reference is a question of law based on subsidiary findings of fact. In re Steed, 802 F.3d 1311, 1316-17 (Fed. Cir. 2015). The issue of reasonable diligence “turns on the factual record, and we review Board determinations as to diligence for support by substantial evidence in the record.” Id. at 1317. Throughout its opinion, the Board repeatedly characterized PST’s burden as requiring proof of Dr. Nezhat’s “continuous exercise of reasonable diligence.” J.A. 16 (emphasis added); see also id. (requiring proof of “continuous and corroborated diligence”); J.A. 18 (“Patent Owner does not provide sufficient evidence to demonstrate ... that Dr. Nezhat continuously exercised reasonable diligence during the entire critical period.”); J.A. 19 (“[T]he evidence, taken as a whole, does not support Patent Owner’s contention that there was a continuous exercise of reasonable diligence during the entire critical period.”); J.A. 22 (“Patent Owner does not provide sufficient evidence to demonstrate Dr. Nezhat’s continuous exercise of reasonable diligence_”). The standard demanded by the Board is too exacting and in conflict with our precedent.
*1009A patent owner need not prove the inventor continuously exercised reasonable diligence throughout the critical period; it must show there was reasonably continuous diligence. See, e.g., Tyco Healthcare Grp. v. Ethicon Endo-Surgery, Inc., 774 F.3d 968, 975 (Fed. Cir. 2014); Monsanto, 261 F.3d at 1370. Under this standard, an inventor is not required to work on reducing his invention to practice every day during the critical period. See Monsanto, 261 F.3d at 1369. And periods of inactivity within the critical period do not automatically vanquish a patent owner’s claim of reasonable diligence. For example, in Monsanto, we held that substantial evidence supported the jury’s presumed finding that an inventor was reasonably diligent where there was no corroborating evidence of any activity for a series of months. Id. at 1370. We explained that, notwithstanding the absence of daily notebook entries and the resulting gaps during the critical period, “the work involved in the experiments was continuous in nature” and therefore the reduction to practice was “reasonably continuous.” Id. In Brown v. Barbacid, we held that the patent owner’s evidence of diligence during a 31-day critical period was “sufficient to show substantially continuing activity” despite the lack of activity during six single-day gaps. 436 F.3d at 1380-83. Our holdings in these cases are consistent with the purpose of the diligence inquiry. In determining whether an invention antedates another, the point of the diligence analysis is not to scour the patent owner’s corroborating evidence in search of intervals of time where the patent owner has failed to substantiate some sort of activity. It is to assure that, in light of the evidence as a whole, “the invention was not abandoned or unreasonably delayed.” Id. at 1379. That an inventor overseeing a study did not record its progress on a daily, weekly, or even monthly basis does not mean the inventor necessarily abandoned his invention or unreasonably delayed it. The same logic applies to the preparation of a patent application: the absence of evidence that an inventor and his attorney revised or discussed the application on a daily basis is alone insufficient to determine that the invention was abandoned or unreasonably delayed. One must weigh the collection of evidence over the entire critical period to make such a determination.
Our decision in In re Mulder, 716 F.2d 1542, 1542-46 (Fed. Cir. 1983), does not instruct otherwise. The Board cites In re Mulder for the proposition that “[e]ven a short period of unexplained inactivity may be sufficient to defeat a claim of diligence.” J.A. 16. In In re Mulder, a competing reference was published just days before the patent at issue was constructively reduced to practice. 716 F.2d at 1544. The patent owner was tasked with showing reasonable diligence during a critical period lasting only two days. Id. at 1545. But the patent owner did not produce any evidence of diligence during the critical period. Id. Nor could it point to any activity during the months between the drafting of the application and the start of the critical period. Id. Although the critical period spanned just two days, we declined to excuse the patent owner’s complete lack of evidence. Id, In re Mulder does not hold that an inventor’s inactivity during a portion of the critical period can, without more, destroy a patent owner’s claim of diligence.
Here, the Board’s erroneously heightened burden of proof infected its analysis. First, the Board did not properly weigh PST’s evidence under a rule of reason. See Price v. Symsek, 988 F.2d 1187, 1195 (Fed. Cir. 1993) (explaining that under a rule of reason analysis, “[a]n evaluation of all per*1010tinent evidence must be made so that a sound determination of the credibility of the inventor’s story may be reached”). Rather than evaluating PST’s evidence as a whole, the Board fixated on the portions of the critical period where PST did not provide evidence of Dr. Nezhat’s specific activities to conclude Dr. Nezhat’s exercise of diligence was not “continuous.” See J.A. 21-22. In doing so, the Board repeatedly condemned PST for not being “sufficiently specific as to facts and dates for the entire critical period during which diligence is required.” J.A. 21; see also id. (noting a portion of the critical period where “Dr. Nezhat does not identify any specific activities undertaken or the dates of those activities”). The Board similarly dismissed Dr. Nezhat’s testimony that he “performed approximately four to six surgeries per week” and worked “roughly 80 hours” per week, J.A, 686, criticizing that Dr. Nezhat was “unable to identify any single date ... on which he was scheduled for a surgery or had another specific conflict that would have prevented him from working on preparing the application for filing.” J.A. 20.
Under a rule of reason analysis, PST was not required to corroborate every day the application was worked on, every surgery Dr. Nezhat performed, or specify precisely what work was done. See In re Jolley, 308 F.3d 1317, 1328 (Fed. Cir. 2002) (“[C]orroboration may be provided by sufficient independent circumstantial evidence, and corroboration of every factual issue contested by the parties is not a requirement of the law.”). Such corroboration is particularly unnecessary when, as here, the record indisputably shows that activities must have occurred within the relevant timeframe. For example, Dr. Ne-zhat testified that, from the time he received the questions in Mr. Heslin’s March 12 letter until the two met on March 16, “most probably we did everything he wanted us to do to go to be prepared for him” and that he was “sure we were pi-epared to go and see him.” J.A. 1316:17-1317:5. The Board criticized this testimony, finding that Dr. Nezhat “does not recall when or how he responded to [Mr. Heslin’s] questions.” J.A. 21. But in this case, such specificity was unnecessary. The record clearly shows that Dr. Nezhat and Mr. Heslin met on March 16, J.A. 604, and Dr. Nezhat provided comments to Mr. Heslin by at least April 13 when Mr. Heslin sent a letter with a revised application to Dr. Nezhat saying he “tried to reflect the comments you made when we last met.” J.A. 607. The revised draft added 7 claims, 8 drawings, and expanded the specification. J.A. 607-31. The parties do not dispute this evidence. With this evidence before it, the Board did not need the specific dates on which Dr. Nezhat provided comments and the content of those comments to evaluate Dr. Nezhat’s testimony under a rule of reason.
The Board compounded its error by summarily dismissing the activities of Dr. Nezhat’s attorney, Mr. Heslin, after determining that PST’s evidence of Dr. Ne-zhat’s activities did not alone establish reasonable diligence. J.A. 22 (“[W]e also need not reach and, therefore, do not address whether Patent Owner provides sufficient evidence to demonstrate Mr. Heslin’s continuous exercise of reasonable diligence for the entire critical period.”). This was error. An attorney’s work in preparing a patent application is evidence of an inventor’s diligence. See Brown, 436 F.3d at 1380 (citing Bey v. Kollonitsch, 806 F.2d 1024, 1030 (Fed. Cir. 1986)). By preparing and filing the application that would issue as the ’384 patent on behalf of Dr. Nezhat, Mr. Heslin was acting as Dr. Nezhat’s agent. The Board was required to weigh Mr. Heslin’s testimony and evidence of activity as an extension of Dr. Nezhat’s. And it was required to weigh this evidence by applica*1011tion of a rule of reason over the course of the entire critical period.
By failing to consider Mr. Heslin’s activities, the Board erroneously found that PST failed to show diligence during periods when Mr. Heslin was working to constructively reduce the invention to practice. Between the March 16 conference and Dr. Nezhat’s receipt of the second draft application on April 13, the Board criticized that “Dr. Nezhat does not identify any specific activities undertaken or the dates of those activities.” J.A. 21. In addition to the work that undoubtedly must have occurred in order to result in a new draft application, the record indicates that during this time, Mr. Heslin at least reviewed and revised the application on April 7. J.A. 604. The Board also flagged PST’s lack of evidence for actions Dr. Nezhat took after receiving the second draft until Mr. Heslin filed the application with the PTO, J.A, 21, but during this time, Mr. Heslin prepared the application and accompanying disclosure statement and filed these documents on May 1. J.A. 604. And Mr. Heslin testified that it was his usual practice to dictate the text of the application and allow time, around eight days, for word processing to transcribe his edits. J.A. 1528 at 33:1-20, Evidence of this sort, where the record indisputably shows an attorney was working to constructively reduce an invention to practice, highlights the importance of considering an attorney’s prosecuting activities as part of the inventor’s diligence.
The Board’s focus on gaps of inactivity also led it to make fact findings unsupported by substantial evidence. The most egregious of these concerns the Board’s finding of inactivity between March 12 and March 15. J.A. 21-22. The record indicates that on March 12, in response to comments received on March 2, Mr. Heslin requested that they “meet to discuss finalizing the draft.” J.A. 927. March 12, 1998 was a Thursday. Mr. Heslin’s billing records confirm that he met with Dr. Nezhat the following Monday, March 16. J.A. 604. The Board determined that the period from Thursday to Sunday constituted a gap of inactivity, finding in support that “[n]o evidence is presented as to the nature of th[e] comments”. Dr. Nezhat sent on March 2 and “[n]o evidence is presented as to the nature of [the March 16] conference.” J.A. 20-21. First, the Board’s finding that there was “no evidence” regarding the March 2 comments is unsupported by the record. Mr. Heslin’s March 12 letter explicitly refers to these comments and explains that Dr. Nezhat’s comments “correctly note that the inclusion of pins will increase the available electrode surface .area relative to conventional (pinless) forceps.” J.A. 927. Second, ample evidence described the “nature” of the March 16 conference. Mr. Heslin’s March 12 letter requesting a meeting asked Dr. Nezhat to clarify whether a device disclosed in another reference “would provide the same increase in' electrode area as your device.” Id, The two met on March 16, which was billed to the patent application matter. J.A. 604. An April 13 letter from Mr. Heslin states, “I have tried to reflect the comments you made when we last met,” which included “various alternative electrode embodiments and-described their benefits both in terms of increased electrode area and enhanced current containment.” J.A. 607. Together, this evidence poses a question regarding the invention, shows the two met on March 16, and shows the question was answered when the two last met. The Board’s finding that this long weekend amounted to a period of inactivity is unreasonable on its face.
The Board concluded that the three gaps collectively demonstrated a lack of diligence during the entire critical period. The Board makes clear that it is “the *1012evidence, taken as a whole,” which causes them to find a lack of diligence “during the entire critical period.” J.A. 19. The Board did not make a fact finding that any one of these gaps would be sufficient to establish a lack of diligence for, the entire critical period. The dissent agrees that the Board’s fact findings regarding two of the three gaps are not supported by substantial evidence. Thus, even under the dissent’s analysis, this case should be remanded for the Board to make the fact finding about whether the evidence presented regarding what it calls Gap 1 would be sufficient alone to find a lack of diligence. The Board’s erroneous finding of inactivity between the periods March 12 to March 15 and April 13 to May 1 requires remand. It is for the Board to reconsider the whole record and make a fact finding on diligence in the first instance.
For these reasons, the Board’s decision that PST failed to antedate JP ’551 is vacated and remanded for proceedings consistent with this opinion. On remand, the question before the Board is whether all of PST’s evidence, considered as a whole and under a rule of reason, collectively corroborates Dr. Nezhat’s testimony that he worked reasonably continuously within the confines of his and Mr. Heslin’s occupations to diligently finalize the patent application during the 81 day critical period.2 If the answer is yes, the Board is to consider PST’s evidence of earlier conception in the first instance.
C. Claim Construction
In IPR proceedings, the Board gives claims them broadest reasonable interpretation consistent with the specification. In re Cuozzo Speed Techs., LLC, 793 F.3d 1268, 1279 (Fed. Cir. 2015), aff'd, — U.S.-, 136 S.Ct. 2131, 195 L.Ed.2d 423 (2016). PST argues the Board erred in its construction of the term “perforated,” which appears in asserted claims 11 and 38: “wherein at least one of the jaws is perforated to permit the release of steam during use.” The Board construed these claims as requiring that “at least one jaw has one or more perforations or passages formed therethrough, so that steam generated between the jaw and tissue is re-leased_” J.A. 10-11 (emphasis added). The Board reasoned that the specification, which states “the jaws may be perforated or otherwise provided with passages,” describes “passages as another form of perforations and does not distinguish between these terms.” J.A. 9 (quoting ’384 patent at 3:25-27). The Board cited extrinsic dictionary definitions in support, including that to “perforate” means to pierce, puncture, or make any hole. J.A. 10. We review the Board’s claim construction de novo except for subsidiary fact findings, which we review for substantial evidence. Cuozzo, 793 F.3d at 1280.
Determining the plain meaning of a term to a skilled artisan at a particular time is a fact finding, and as such we review it for substantial evidence. Teva Pharm. USA, Inc. v. Sandoz, Inc., — U.S.-, 135 S.Ct. 831, 843, — L.Ed.2d -(2015). In this case, the Board concluded that the plain meaning of the term “perforated” included passages. It referenced extrinsic evidence, namely dictionary definitions which supported its determination of the plain meaning. The legal part of claim construction is the determination of the meaning of the term in the *1013claim in light of the patent’s intrinsic record. PST argues the description in the specification that the device “may be perforated or otherwise provided with passages” evidences a difference in meaning between passages and perforations. PST ,Br. 47 (quoting ’384 patent at 3:25-27). We agree. The patentee is free to define or limit a term used in a patent. See Thorner v. Sony Comput. Entm’t Am. LLC, 669 F.3d 1362, 1365-66 (Fed. Cir. 2012). We conclude that the specification’s separation of the terms perforated and passages with the disjunctive phrase “or otherwise” makes clear that the patentee intended that the term “perforated” is not the same as “passages.” The patentee claimed only jaws that are “perforated”; this claim does not extend to passages. In light of the intrinsic record, we conclude that the term “perforated” is not coextensive with or the same as “passages.” Thus the Board’s finding that JP ’551 disclosed at least one “perforated” jaw because JP ’551 referenced a passage cannot be supported. We vacate the Board’s decision invalidating claims 11, 38, 41-44, 46, 47, and 49 over JP ’551 and remand for proceedings consistent with this construction.
Conclusion
For the foregoing reasons, we hold the Board’s acceptance of Olympus’ submission of JP ’551 was harmless, vacate and remand the Board’s determination that PST failed to antedate JP ’551, and vacate the Board’s decision invalidating claims 11, 38, 41-44, 46, 47, and 49 over JP ’551 and remand for proceedings consistent with our construction of the term “perforated.”
VACATED AND REMANDED
Costs
Costs to PST.
. Because the Board determined that PST had not proven reasonable diligence, it did not reach the issue of conception. We decline to consider PST's evidence of conception for the first time on appeal and therefore limit our opinion to the issue of reasonable diligence.
. This evaluation must include: (1) Mr. Hes-lin’s testimony of his and Dr. Nezhat's activities; (2) Mr, Heslin’s billing records identifying work on March 4, March 16, April 7, and May 1; (3) Mr, Heslin's March 12 letter thanking Dr. Nezhat for his March 2 comments and requesting guidance; (4) Mr. Heslin’s April 13 letter and revisions, adding text to the specification, 7 claims, and 8 drawings; and (5) the finalized patent application filed on May 1.