Tuesday, 20 April 2010

T 1281/06 – Cack-handed Lexicography

The present decision is of interest to writers of patent applications who act as their own lexicographers by defining some of the claimed terms in the application.

[...] The reference in independent claims 1 and 5 to a current color mapping function having upper and lower knee points is not clear because the term “knee point” does not have any generally recognised meaning in the field of image processing.

The Board does not accept, in this respect, the appellant’s argument that a definition of this term in the description suffices to render the definition of the claimed subject-matter clear; as a general principle, the meaning of the terms of a claim should be clear from the wording of the claim alone.

As already stated in a previous decision of the present Board (in a different composition), “The Board is aware of the jurisprudence which acknowledges that an exception to this principle may exist in situations in which a patent description would provide unambiguous definitions of certain terms and would also make clear that such definitions apply throughout the complete application, so that, when interpreting the wording of claims, the patent specification would constitute its own dictionary” (cf. T 1271/05 [4.3]).

The Board observes, however, that these conditions are not met under the present circumstances since the current description does not contain any statement which would imply that the term “knee point” should be given a specific meaning applying throughout the entire application.

The passage on page 9, line 23-33 of the original application, referred to by the appellant in the statement of grounds, merely refers, in association with Figure 3, to one particular illustration of a color mapping function having upper and lower knee points. It fails, however, as such, to give a clear statement which could be identified as a definition.

The following passages of the description on page 9, line 33 to page 10, line 34, admittedly, provide additional information as to the parameters which enable a determination of said knee points.

Similarly, the passage on page 3, line 30 to page 4, line 32, suggests that the knee points constitute specific points of the color mapping function fulfilling certain conditions.

In addition, the indications in dependent claims 2 and 6 regarding the form of the color mapping function could be regarded as an alternative definition of the concept of “knee points” within the meaning of the present application.

However, the description does not specify whether any of these various indications should apply to the entire application.

Instead, the passage on page 11, lines 12- 18, also cited by the appellant, discloses that the positions of the knee points can, alternatively, be determined by the host computer based on user selected inputs such as, for example, the system color flow gain setting or the dynamic range setting. This non-exhaustive list of parameters suggests in effect that any parameter indirectly relating to the Doppler imaging data could be validly relied on to define knee points.

The passage on page 12, lines 4-32, adds to the confusion in that it specifies that the knee points could be moved by the user through the operator interface by moving a graphical symbol to a selected position. In the absence of any limitation as to these selected positions, this embodiment suggests that the knee points can be selected in an essentially arbitrary manner, i.e. even without consideration of the parameters previously referred to.

As a consequence, the Board is unable to recognize a unique and unambiguous definition of the term “knee point” in the description which would also apply to the claims. It is hence not possible, in line with decision T 1271/05, to associate to the features of the lower and upper knee points recited in independent claims 1 and 5 any specific meaning which would distinguish said points from any point defined by the color mapping function. [2.1]

The aspect regarding the definition of the knee points is considered essential under the present circumstances in view of the technical problem which the claimed invention intends to solve. As indicated on page 3, lines 17-28, of the original application, the mapping functions known from the prior art "do not allow the user or the system to optimize the image information".

The solution of this problem resides in the selection of a color mapping function having the optimal form […], which form presents a segment of increased slope in order to better visualize the actual distribution of power densities for a predetermined range of Doppler flow powers.

This finding implies that essential information for the definition of the claimed system and method is lacking from independent claims 1 and 5. More specifically, the information as to the criteria to be used for establishing the optimal form and to determine the analyzing steps to be made are missing from independent claims 1 and 5, respectively. [2.2]

[…] For the reasons developed above, the Board concludes that the requirements of A 84 EPC 1973 as to clarity are not met. [2.4]

So when you define terms in the application, be sure to make clear that the definitions concern the application as a whole.

To read the whole decision, click here.


Rimbaud said...

Interesting decision.
As often, it's kind of you, Oliver, to derive practical tips therefrom.

Anonymous said...

It seems that for the purpose of clarity, the patent description can be its own dictionary only if it would not be a problem anyway to introduce the definition into the claim (in the sense of Art. 123(2) EPC). So the practical benefit of this exception seems very limited.

Trying to find the origin of this exception, I have so far found T 311/93, which indeed states that a specification can be its own dictionary. However, that was in the context of interpreting the priority document, NOT for the purpose of clarity. I think it's quite normal to use the description for interpreting the claims, e.g. to show that a claim term does cover a feature disclosed in a prior art document or for the purpose of infringement, but the point of Art. 84 is that the claims should be clear on their own, i.e. without the description (T 0988/02, pt. 3.3.1).

Clearly there is case law applying "a specification can be its own dictionary" to clarity, but there is also case law applying Art. 69 EPC to clarity. I thought the latter approach is now rejected by most Boards. I wonder what most Boards think of the former.

T 1321/04 (mentioned in the Case Law book) does not refer to Art. 84. T 523/00 (referred to by T 1321/04) does state that the specification can be its own dictionary in the context of clarity, but derives this from Art. 69, and unless I'm very mistaken most Boards do not agree that Art. 69 is to be used when examining clarity.

So far I'm not convinced that there is much case law to be found in support of this principle that claim terms may be defined in the description.