04/02/2025
Described as “game changing”, researchers at the University of Cambridge have developed an artificial intelligence (AI) tool named IRON (Integrated Radiogenomics for Ovarian Neoadjuvant therapy) to predict ovarian cancer patients’ responses to chemotherapy. By integrating diverse patient data – including CT images, blood tests, genomic results, and health records – IRON enables clinicians to make personalised treatment decisions more efficiently.
From a patent perspective, it is interesting to see how such a therapy may not be considered so game-changing.
IRON appears to have been initially filed as a UK patent application in June 2020. An international patent application was filed a year later claiming priority to this UK patent application. Subsequently, this international patent application appears to have entered the national/regional phase in the US, China, Japan, Australia, Canada and European Patent Offices. We’ve looked at the public files for the US and European cases, which are still pending more than four years after the initial patent application was filed.
At both the US and European Patent Offices, objections have been raised to novelty and indicated that the subject-matter of the application may not be patent eligible. In other words, examiners at both patent offices have taken the view that the technology was already in the public domain before the patent application was filed and, furthermore, if there is any difference over the technology that was already in the public domain, this difference isn’t technical and therefore not eligible for patent protection. For an AI patent application, this is the kind of objection that may be raised if the AI technology is already known and the only difference is the application of the AI or the data on which the AI is trained.
Can anything be done about this?
Speed: While more than four years from an initial application and to still be arguing with patent office examiners may seem slow, it’s quite usual. A different filing strategy and strategic use of acceleration tracks often available at patent offices can significantly speed things up if required such as to open a licencing revenue stream, have an enforceable right in the courts, or most commonly appear more attractive to potential investors. However, there are many commercial reasons why slow is good such as to delay costs, to keep third-parties guessing the final scope of protection, or to wait and see the success of a commercial embodiment.
Patentability: In relation to the problems of patentability, sometimes nothing can be done. An invention is inherently unpatentable and as we say above, in the AI space, while a tool may be considered game changing practically, the AI technology may already be known and the only difference is the application or the data on which the AI is trained. However, sometimes the drafting and prosecution of the patent application can help, and life sciences applications of AI can be particularly difficult. It is rare that a patent professional would be knowledgeable in both AI and life sciences. At Reddie & Grose, we are fortunate to have a team member who is well-versed in both fields to post-graduate level, and we also collaborate as a team, with drafters from different tech or life science backgrounds. This approach ensures that we can navigate the complexities of patentability where AI and life sciences intersect, and provide tailored strategies for inventors to maximise their potential of securing robust intellectual property protection.
This article is for general information only. Its content is not a statement of the law on any subject and does not constitute advice. Please contact Reddie & Grose LLP for advice before taking any action in reliance on it.
