EPO Bioinformatics Roadshow

There is still a relatively prevalent view in the field of bioinformatics and medical informatics that software-based inventions are not patentable at the EPO. A large proportion of the innovation that is being made in this field is primarily or entirely implemented in software and does not involve new hardware. As such, many applicants miss out on potentially valuable protection because of the false belief that they would be wasting their time (and money) by filing at the EPO or talking to a European Patent Attorney.  This could not be further from the truth, and the EPO is actually a comparatively favourable jurisdiction for computer-implemented inventions (CIIs) in the field of bioinformatics and medical informatics.

On October 21-23 rd, we led a pair of EPO Examiners on a mini-tour of the UK to meet applicants and attorneys working in the field of bioinformatics and medical informatics. The aim of this series of seminars was for the EPO team in charge of examining applications in this field to improve the understanding of how they assess CIIs, with a view to increase the number of applications filed, and their grant rate.

The seminars focused on the infamous “2 hurdles” approach used by the EPO to assess “mixed inventions” – i.e. inventions that involve a combination of technical and non-technical features (in particular, inventions involving mathematical methods steps and/or features related to the presentation of information).

The first hurdle stems from Article 52 EPC (excluded subject-matter) and tests whether the claimed subject-matter relates to excluded subject-matter “as such”. This is assessed by separating a claim into each of its individual features, and evaluating whether at least one of these features, when considered in isolation, is technical. The presence of a computer, memory, display, physical input means, or any technical step is sufficient to overcome this hurdle. For example, when claiming a method comprising mathematical steps (a)-(c), the inclusion of the words “computer-implemented [method]” is sufficient to overcome this hurdle.

The second hurdle stems from Article 56 EPC (inventive step). It is established practice at the EPO that only those features that contribute to the technical character of the invention are taken into account for the purpose of assessing the inventiveness of a claim. In practice, this means that examiners have to separate out those features that cannot be seen to contribute to the technical character of the invention, and assess inventiveness exclusively based on the remaining (technical) features. For example, when claiming a computer-implemented method comprising mathematical steps (a)-(c), if mathematical steps (a)-(c) are deemed not to contribute to the technical character of the invention, then the Examiner is left with the feature of “computer-implemented method [of doing something]”, and the closest prior art for this can be any computer-implemented method.

However, features which, when taken in isolation are non-technical can nonetheless, in the context of the invention, contribute to producing a technical effect serving a technical purpose. Any feature for which this can be successfully demonstrated will be taken into account in the assessment of inventive step. The skill therefore is in ensuring that features related to mathematical methods and/or the presentation of information are specified with sufficient detail, credibly and causally linked to a technical purpose, such that they are taken into account in this assessment. Specific examples of when this is and isn’t successful were discussed, as well as common pitfalls to avoid getting stuck between a rock (lack of inventive step due to insufficient specification of technical features and/or technical purpose) and a hard place (added matter). Watch out for more on this in a subsequent post…