T 1896/11 25-11-2016
Téléchargement et informations complémentaires:
A SYSTEM AND METHOD FOR DETECTING SIGNAL ARTIFACTS
Amendments - broadening of claim (yes)
Amendments - added subject-matter (yes)
Summary of Facts and Submissions
I. The appeal lies from the decision of the opposition division, dispatched on 28 June 2011, to revoke the European patent number EP 1 658 578.
Opposition was filed against the patent as a whole and based on:
Article 100(a) EPC 1973 (Article 54(1),(2) EPC 1973 and Article 56 EPC 1973),
Article 100(b) EPC 1973 (Article 83 EPC 1973),
Article 100(c) EPC 1973 (Article 123(2) EPC).
During opposition proceedings, an objection under Article 123(3) EPC was raised.
The opposition division held that the patent as amended during the opposition proceedings did not comply with Article 123(3) EPC.
II. In the notice of appeal dated 29 August 2011, the appellant (patent proprietor) requested that the decision under appeal be set aside and the patent be maintained as granted.
Oral proceedings were requested as an auxiliary measure.
In the statement setting out the grounds of appeal dated 7 November 2011, the appellant requested that the patent be maintained as amended on the basis of claims corresponding to a main request or, alternatively, on the basis of claims corresponding to a first auxiliary request, the text of the both claim sets being filed with the statement of grounds.
The appellant also requested "to hear an opinion of an expert appointed by the patent proprietor" with regard to the interpretation of the claimed subject-matter (Article 117 EPC and Rule 117 EPC).
III. In reply, the respondent (opponent) requested that the appeal be dismissed.
Oral proceedings were requested as an auxiliary measure.
IV. The Board issued a communication pursuant to Article 15(1) RPBA in preparation of oral proceedings, setting out the issues to be discussed, notably Article 123(3) EPC and the relevance of Article 69 EPC in determining the extent of protection in the present case. It was also indicated that the clarity of amendments (Article 84 EPC 1973), sufficiency of disclosure (Article 83 EPC 1973) and added matter (Article 123(2) EPC) may have to be considered.
The Board also envisaged the possibility to remit the case to the opposition division for consideration of novelty (Article 54(1),(2) EPC 1973) and inventive step (Article 56 EPC 1973), where appropriate.
V. Both parties filed additional comments in reaction to the Board's communication.
Moreover, the appellant filed sets of amended claims forming the basis of new requests.
VI. Oral proceedings were held on 25 November 2016.
VII. The final requests of the appellant were that the decision under appeal be set aside and a patent be maintained in amended form according to one of the following sets of claims:
Main request:
Claims 1 to 8 filed as a main request with the letter of 20 June 2016;
First auxiliary request:
Claims 1 to 8 filed as a first auxiliary request with the letter of 1 July 2016;
Second auxiliary request:
Claims 1 to 8 filed as a main request with the statement setting out the grounds of appeal dated 7 November 2011;
Third auxiliary request:
Claims 1 to 8 filed as an auxiliary request with the statement setting out the grounds of appeal dated 7 November 2011.
The request to hear the opinion of an expert was withdrawn.
VIII. The final request of the respondent was that the appeal be dismissed.
IX. Independent method claim 5 of each of the appellant's respective requests reads as follows. The wording of the other claims does not play a role in the present decision and so is not reproduced here.
Main request
"A method for detecting in a patient monitoring system (10) a signal artifact in a monitored data signal (11), the method comprising the steps of:
- receiving at least two monitored data signals (11),
- determining a first cross-correlation matrix (rGlobal) for the at least two monitored data signals (11) received over a first period of time,
- determining second cross-correlation matrices (rlocal,i, i=1,...,N) for the at least two monitored data signals (11) received over N second periods of time, said second periods of time being shorter than the first period of time,
- determining N deviations (Di, i=1,...,N) between the second cross-correlation matrices (rlocal,i) and the first cross-correlation matrix (rGlobal),
- determining an average deviation (Daverage) from the N deviations (Di), and
- determining, based upon variations of the deviations (Di) from the average deviation (Daverage), whether one of the at least two monitored data signals (11) received over the second periods of time has been corrupted with an artifact."
First auxiliary request
With respect to claim 5 of the main request, only the final method step has been amended and now reads:
"- determining whether one of the at least two monitored data signals (11) received over any of the second periods of time has been corrupted with an artifact based upon a variation of the determined deviation (Di) between the second cross-correlation matrix determined for the at least two monitored data signals (11) received over the respective one of the second period of time and the first cross-correlation matrix from the average deviation (Daverage)."
Second auxiliary request
With respect to claim 5 of the main request, only the final two method steps have been amended and now read:
"- determining an average deviation (Daverage) from the deviations (Di),,[sic] and
- determining based upon the variation in the deviation associated with the variation of the local correlation matrix from the global correlation matrix away from the average deviation (Daverage), whether one of the at least two monitored data signals (11) received over the second periods of time has been corrupted with an artifact."
Third auxiliary request
With respect to claim 5 of the main request, only the final two method steps have been amended and now read:
"- determining an average deviation (Daverage) from the deviations (Di),,[sic] and
- determining based upon the variation in the average deviation (Daverage), whether one of the at least two monitored data signals (11) received over the second periods of time has been corrupted with an artifact."
X. The arguments of the appellant with respect to the main request, insofar as they are relevant to the present decision, may be summarised as follows.
Following a first line of argument, when reading claim 5 of the granted patent in isolation (i.e. without reference to the description) the skilled person would have serious doubts about how the claim should be interpreted. In particular, in the context of the claim, it would not be clear to the skilled person how any variation in the average deviation could be derived. Claim 5 of the granted patent defined merely that a first correlation matrix and a second correlation matrix were determined, that a deviation between the first and second correlation matrices was determined and that an average deviation was determined from the deviation recorded over multiple past periods. No indication was contained in the claim to suggest that multiple average deviations should be calculated. Indeed the use of the singular "an average" suggested the contrary. Moreover, even if the skilled person could read into the wording that the average deviation was to be updated each time a second cross-correlation matrix became available (as suggested by the respondent) the skilled person would realise that monitoring any changes in the average deviation would give, at best, only a very "blurred" indication of changes in the deviation from one time period to the next.
In view of the doubts concerning the meaning of claim 5, the skilled person would have to turn to the description to gain a better understanding of what was meant by the claimed subject-matter.
From the description, the skilled person would see from equation 6 that an average deviation Daverage was calculated from N deviations Di. A signal was identified as being corrupted by an artifact during one of the N second periods of time when its "[associated] deviation (equation 5) varies largely from the average deviation" (paragraph [0032]). In other words, a large variation of Di from Daverage was indicative of an artifact. This was illustrated in Figure 3 which showed the deviation away from the average deviation, whereby the average deviation was depicted by the horizontal axis. The suggestion in paragraph [0035] that "the red dashed line" represented the average deviation was incorrect. The description makes no reference to a variation in the average deviation.
The skilled person would therefore realise that claim 5 of the granted patent contained an obvious mistake in the form of an inaccurate technical statement which was evidently inconsistent with the totality of the disclosure of the contested patent. Claim 5 of the granted patent did not reflect the procedure laid out in the description for determining the presence of an artifact. Following decision T 108/91 (see Reasons, point 2.3, 7th and 8th paragraphs), the offending method step could be replaced with an accurate statement of the step concerned without infringing Article 123(3) EPC.
Following a second line of argument, the appellant noted that Article 123(3) EPC stipulated that "The European patent may not be amended in such a way as to extend the protection it confers" (emphasis added). In order to establish the protection conferred by claim 5 of the granted patent, and thus to identify the boundary of acceptable amendments, Article 69 EPC had to be taken into account. Specifically, in accordance with Article 69 EPC, the extent of protection was a result of the interpretation of the claim, in the light of the description and drawings. In the current case, it was in fact essential to apply Article 69 EPC, since the skilled person would have serious doubts about the meaning of the claim when reading it in isolation (as shown above) and would have to rely upon the teaching of the description in order to understand the claim. The description set out that it was the variation of the deviation Di from the average deviation Daverage which was used in the determination of artifacts. This interpretation differed from the literal wording of claim 5 of the granted patent but, taking Article 69 EPC into account, the protection conferred by claim 5 necessarily extended to this interpretation. Since the current amendments to claim 5 reflected this understanding, they did not extend the protection conferred and therefore did not infringe Article 123(3) EPC.
So on two counts, the appellant concluded that the amendments did not infringe Article 123(3) EPC.
XI. The arguments of the respondent with respect to the appellant's main request, insofar as they are relevant to the present decision, may be summarised as follows.
In claim 5 of the granted patent, the artifact detection was based on the variation in the average deviation Daverage. In amended claim 5, the artifact detection was based on a comparison of the average deviation Daverage with the deviations Di between the first and second cross-correlation matrices. The comparison performed in claim 5 of the granted patent had been replaced by a different comparison. The scope of protection conferred by claim 5 had therefore been altered. Consequently, the amendments were not allowable under Article 123(3) EPC.
The wording of claim 5 of the granted patent was clear and could be understood when read in isolation, i.e. without reference to the description. Specifically, claim 5 of the granted patent stated that the deviation between the first and second cross-correlation matrices was recorded over multiple past periods of time. This meant that for each second period of time, an additional deviation Di would be calculated, giving rise to a plurality of deviations. As the latest deviations Di were each fed into the calculation of the average deviation Daverage, some fluctuation of the average deviation would occur with respect to the previous average. The skilled person would understand from claim 5 that it was this fluctuation of the average deviation Daverage which gave an indication of the presence of an artifact.
Whilst the averaging operation would inevitably give rise to some "blurring" of the variation, the extent of the "blurring" would not necessarily disguise the variation. In particular, there was no indication in claim 5 of the granted patent of how many samples were used to calculate the average. Any "blurring" could be minimised by minimising the number of samples involved.
Since claim 5 of the granted patent could be meaningfully interpreted in this sense, it did not contain an obvious mistake in the form of an inaccurate technical statement. There was therefore no reason to amend the final method step of claim 5 of the granted patent in a manner which changed the meaning thereof and no reason to turn to the description to understand the claim.
Notwithstanding the fact that there was no justification in seeking reinterpretation of the claims by consulting the description, even if the skilled reader would have recognised that the claims appeared to differ from what was explained in the description, the application as filed also failed to provide for a clear and unambiguous basis for the "correction" suggested by the appellant. Notably, original claim 7 recited the step of determining the presence of an artifact "based upon the average deviation".
Moreover, the explanation of the occurrence of alarm signals given in paragraphs [0032] to [0035] of the patent did not appear to correlate with what was illustrated in Figure 3. In particular, paragraph [0033] explained that an alarm having "a large deviation away from Daverage" was likely to be a false alarm. However, paragraph [0035] indicated that the alarm having the largest variation away from the red dashed line depicting the average deviation in Figure 3 (i.e. the first alarm) was in fact a true alarm and the alarm having the smallest variation away from the average deviation in Figure 3 (i.e. the fourth alarm) was in fact a false alarm. Paragraphs [0032] to [0035] were therefore contradictory and could not be used to derive the allegedly intended meaning.
Consequently, the respondent concluded that the amendments to claim 5 violated Article 123(3) EPC.
Reasons for the Decision
1. The appeal is admissible.
2. Appellant's main request
2.1 Admissibility
The claims of the appellant's main request were filed with the letter of 20 June 2016 and thus after the statement setting out the grounds of appeal had been filed.
The respondent did not object to the admissibility of the main request. The Board also had no reason to question it considering that said letter was a reaction to the Board's communication of 6 June 2016.
The appellant's main request was therefore admitted into the proceedings (Article 13(1) RPBA).
2.2 Article 123(3) EPC
2.2.1 The final step of method claim 5 of the patent as granted reads "determining based upon the variation in the average deviation (Daverage) whether one of the at least two monitored data signals (11) received over the second period of time has been corrupted with an artifact" (emphasis added by the Board).
The final step of method claim 5 of the appellant's main request reads "determining, based upon variations of the deviations (Di) from the average deviation (Daverage) whether one of the at least two monitored data signals (11) received over the second periods of time has been corrupted with an artifact" (emphasis added by the Board).
2.2.2 The appellant submitted that claim 5 of the granted patent contained an inaccurate technical statement. From the description it was clear what was obviously intended and, following decision T 108/91 (OJ 1994, 228), claim 5 could be corrected to reflect this intention.
2.2.3 The Board notes that claim 5 of the granted patent sets out that an average deviation (Daverage) is determined from the deviation (D) "recorded over multiple past periods". This gives the reader the clear teaching that a plurality of deviations are recorded. In particular, first and second cross-correlation matrices are derived over a number of different first and second periods of time respectively and that the corresponding deviations between these first and second cross-correlation matrices are determined accordingly. From the resulting plurality of deviations recorded in this manner, an average deviation can then be determined. In view of the fact that claim 5 of the patent as granted also refers to the "variation in the average deviation", the reader would understand that a number of separate average deviations must be somehow determined in order to establish a variation therein. Exactly which mechanism is used to obtain a plurality of average deviations is not defined in claim 5. The claim provides no details of which of the recorded deviations are to be fed into the calculation of the average deviation. However, the absence of such details does not affect the understanding of the claim in this respect. Irrespective of how the average is determined - for example, the average of all recorded deviations could be calculated and updated each time a new second cross-correlation matrix is determined or a sliding average of the last ten deviations could be calculated each time a new second cross-correlation matrix is determined - the skilled person would understand from the claim that (i) a plurality of average deviations has to be determined and (ii) any change in these averages has to be monitored. The Board therefore cannot agree with the appellant that the skilled person would not understand at all from claim 5 of the patent as granted how a variation in the average deviation could be determined.
As the appellant submitted, performing an averaging operation would dilute any fluctuation from one individual deviation (Di) to the next (Di+1) and consequently make any such fluctuations harder to detect. Nevertheless, some variation of the averages (Daverage) will be apparent. Although this may seem an unusual way of identifying a variation, it is nonetheless plausible.
The Board therefore agrees with the respondent's view that the skilled person would understand claim 5 of the granted patent as it stands and would not have any reason to suspect that it contains an inaccurate technical statement.
The Board does not contest that the teaching of the description of the patent as granted does not coincide with what is defined in claim 5. Nevertheless, the wording of claim 5 is not devoid of meaning and, as shown above, may be interpreted in a plausible manner without recourse to the description. Thus, in contrast to the situation in T 108/91, although an inconsistency between granted claim 5 and the description exists, it is not immediately apparent that what is defined in claim 5 could not be that for which protection was sought.
Moreover, the Board notes that the decision T 108/91 is older than the decision G 1/93 (OJ 1994, 541) in which the Enlarged Board of Appeal dealt with a similar point of law. In G 1/93, the Enlarged Board had to decide upon the case in which a technical feature, which was not disclosed in the application as originally filed and which limited the scope of protection of the claims of the granted patent as compared to the application as filed, had been added during examination (see Reasons, point 12). The limiting feature could not be maintained in the patent in view of Article 100(c) EPC, nor could it be removed from the claims without violating Article 123(3) EPC. The Enlarged Board held that "Only if the added feature can be replaced by another feature disclosed in the application as filed without violating Article 123(3) EPC, the patent can be maintained (in amended form)" (see Reasons, point 13, emphasis added by the Board).
Thus, in accordance with G 1/93, it is not allowable to replace a technical feature of a patent claim with another technical feature which causes the claim to extend to subject-matter which was not encompassed by the granted claim. As pointed out in a more recent decision T 195/09 (unpublished), "In this respect decision T 108/91 has been clearly overruled by G 1/93" (see Reasons, point 2.1.5).
In claim 5 of the patent as granted, the artifact detection is based on the variation in the average deviation Daverage. In amended claim 5, the artifact detection is based on the variations of the deviations Di from the average deviation Daverage. According to claim 5 of the granted patent it is therefore necessary to monitor changes in the average deviation. According to claim 5 as amended, it is necessary to monitor changes in the difference between the individual deviations Di and the average deviation Daverage. Thus, the parameter to be monitored in claim 5 of the patent as granted has been replaced by a different parameter and the scope of protection conferred by amended claim 5 had therefore been altered. Consequently, the amendments are not allowable under Article 123(3) EPC.
2.2.4 The appellant also submitted that, taking Article 69 EPC into account, the description had to be used to interpret the claims. The protection conferred by claim 5 of the patent as granted necessarily extended to what was was stated in the description. Claim 5 could therefore be corrected to reflect the description without infringing Article 123(3) EPC.
2.2.5 The Board considers that although the Protocol on the Interpretation of Article 69 EPC makes clear that the extent of protection conferred by a European patent is not to be understood as that defined by the strict, literal meaning of the claims, it does not provide a basis for ignoring the wording of a claim of a granted patent. Indeed Article 1 of the Protocol states that the aim of Article 69 EPC is to combine "a fair protection for the patent proprietor with a reasonable degree of legal certainty for third parties". The Protocol attempts, on the one hand, to ensure that a claim is not interpreted too literally such that a patent proprietor may contend for a (justified) broader interpretation of the claim than its wording would warrant. On the other hand, the Protocol attempts to balance this against the interests of third parties who rely on the wording of the granted claims. In this respect, the second sentence of Article 1 of the Protocol also states that Article 69 EPC should not "be taken to mean that the claims serve only as a guideline and that the actual protection conferred may extend to what, from a consideration of the description and drawings by a person skilled on the art, the patent proprietor has contemplated."
Thus, the Board holds that, in the present case, the description cannot be used to give a different meaning to a claimed method step which in itself imparts a clear, credible, technical teaching to the skilled reader. Otherwise third parties could not rely on what the claim actually states.
In order to determine whether the claimed method step in itself imparts a clear, credible technical teaching, it is necessary to examine "whether (a) the step as claimed is in itself meaningful and plausible from a technical point of view, and (b) there is, prima facie, any inherent incompatibility with the remaining features of the claim" (see decision T 1202/07, unpublished, Reasons, point 2.5).
As has been shown above, the Board considers the step of "determining based upon the variation in the average deviation (Daverage) whether one of the at least two monitored data signals (11) received over the second period of time has been corrupted with an artifact" to be meaningful and plausible from a technical point of view. Moreover, understanding the claim to mean that a variation in the average deviation is used to detect the artifacts is not incompatible with the remaining features of the claim which merely define the steps involved in determining the average deviation.
Consequently, even although claim 5 of the granted patent does not align with the teaching of the description, there is no reason to read a different technical meaning into the term "based upon the variation in the average deviation". To change the interpretation of claim 5 - the wording of which, in isolation, imparts a clear and credible technical teaching - to an entirely different interpretation which could only be derived from the description, would deprive the terms of the claim of any meaning and seriously compromise the legal certainty associated with the granted wording.
2.2.6 In view of the above considerations, the Board concludes that any attempt to redefine the invention to cover something which was not encompassed by the claims of the patent as granted cannot be allowed under Article 123(3) EPC. In the specific case at hand, claim 5 according to the appellant's main request has been amended to define that the presence of an artifact is based on variations in the deviations Di from the average deviation Daverage whereas claim 5 of the patent as granted defined that the presence of an artifact is based on the variation in the average deviation Daverage. Since the amended claim relates to a different concept which did not fall under the wording of claim 5 of the patent as granted, the amendment of claim 5 extends the protection conferred by the granted patent.
2.2.7 Claim 5 of the appellant's main request therefore does not meet the requirements of Article 123(3) EPC.
3. Appellant's first auxiliary request
3.1 Admissibility
3.1.1 The claims of the first auxiliary request were filed with the letter of 1 July 2016 and thus after the statement setting out the grounds of appeal.
3.1.2 In accordance with Article 13(1) RPBA, any amendment to a party's case after it has filed its grounds of appeal may be admitted and considered at the Board's discretion.
One of the criteria applied by the boards when considering the admissibility of late-filed requests is the question of whether the amended claims are prima facie allowable. It must be immediately apparent to the Board that the amended claims constitute a promising attempt to counter all outstanding objections without giving rise to new ones (see Case Law of the Boards of Appeal of the European Patent Office, 8th Edition 2016, Chapter IV.E.4.4.2).
3.1.3 In the present case, as indicated by the respondent, independent method claim 5 suffers from the same deficiency as claim 5 of the appellant's main request. Even the appellant conceded that, when interpreting claim 5 in the same way as claim 5 of the main request was interpreted, the current request obviously did not overcome the objection raised against the main request under Article 123(3) EPC.
3.1.4 The appellant's first auxiliary request is therefore not admissible (Article 13(1) RPBA).
4. Appellant's second auxiliary request
4.1 Admissibility
4.1.1 The claims of the appellant's second auxiliary request were submitted as a main request with the statement setting out the grounds of appeal and are therefore admissible (Article 12(1)(a) RPBA).
4.2 Article 123(3) EPC
4.2.1 Independent method claim 5 infringes Article 123(3) EPC for the same reasons as claim 5 of the present main request.
4.2.2 The appellant did not contest this finding.
4.2.3 The appellant's second auxiliary request is therefore not allowable.
5. Appellant's third auxiliary request
5.1 Admissibility
5.1.1 The claims of the appellant's third auxiliary request were submitted as an auxiliary request with the statement setting out the grounds of appeal and are therefore admissible (Article 12(1)(a) RPBA).
5.2 Article 123(3) EPC
5.2.1 The wording of method claim 5 adheres very closely to the wording of claim 5 of the patent as granted. The final method step essentially returns to the granted wording and now reads "determining based upon the variation in the average deviation (Daverage) whether one of the at least two monitored data signals (11) received over the second periods of time has been corrupted with an artifact", the only difference being the reference to "second periods of time" (emphasis added).
In view of the near identity of wording, the objection raised under Article 123(3) EPC against the main request and the second auxiliary request does not apply to claim 5 of the third auxiliary request.
5.3 Article 123(2) EPC
5.3.1 The appellant submitted that the basis for the amendments vis-à-vis the originally filed claims was to be found in the first two paragraphs of page 7 of the originally-filed application. This passage referred to "a large deviation away from Daverage", implying a variation between one value of the average deviation and another value thereof.
5.3.2 The respondent did not comment on this point but questioned the basis for the plurality of "second cross-correlation matrices" (emphasis added).
5.3.3 The Board notes that the passage referred to by the appellant explains that when any of the monitored signals are affected by an artifact, the deviation between the local correlation matrix rlocal i varies largely from the global correlation matrix rGlobal and the associated deviation Di (rGlobal - rlocal i) varies largely from the average deviation Daverage. In this case, "an alarm is present with a large deviation away from Daverage". This passage does not refer to a variation in the average deviation.
Moreover, original claim 7 contained the step of "determining whether an artifact was detected in one of the at least two event signals based upon the average deviation". This also provides no basis for determining the signal corruption "based upon the variation in the average deviation" (emphasis added), as is currently set out on claim 5.
5.3.4 No other basis for the amendment to claim 5 could be identified in the originally-filed application documents. For this reason, claim 5 does not comply with the requirements of Article 123(2) EPC.
5.3.5 The appellant's third auxiliary request is therefore not allowable.
Order
For these reasons it is decided that:
The appeal is dismissed.