AI in health technology: takeaways from our half-hour webinar on the power of patents
Patent Power Half Hour is the supercharged, information-rich webinar series from Kilburn & Strode. We dive into typical EPO pitfalls and share practical tips on how to avoid them. All delivered in 30 minutes or less!
Here’s a recap of the key takeaways from the webinar that took place on December 8, 2021, covering AI in healthcare. This webinar was a replay of our July webinar on MedTech Patents in Europe (Part 2).
Patenting AI innovation in MedTech
AI and machine learning have established themselves as essential tools in all areas of technology, and healthcare is no exception. In our condensed and information-rich Patent Power Half Hour, we shed the spotlight on two important topics: AI in drug discovery (whether or not you should patent AI innovation in this area) and criteria for obtaining patent protection at the EPO for AI Innovation Medtech (can you get patents in this area).
The answer to the first question may be, and the answer to the second question is clearly yes, if you can comply with the EPO’s examination guidelines. Read more below, but remember that while the EPO’s approach in this area is fairly established, there is always room to push the boundaries, so if you are considering borderline cases you can or maybe even should patent, enter contact us to discuss.
AI and drug discovery
In drug discovery, innovators have a strategic choice. Should you patent your AI-based drug discovery method? Or should you patent the product resulting from this method? Or both? There are a number of strategic advantages to focusing your patenting activity on pharmaceuticals, rather than the tools you use to identify them.
Are you sure you want to disclose your AI method; maybe it is better to protect it as a trade secret? However, you may need to disclose your product on the way to regulatory approval.
AI algorithms are often developed so quickly that the patent system is not ideal for protecting them; the patent application could become redundant too quickly. In contrast, your product is not as likely to change.
Even if you get a complaint for your AI method, how will you know someone is breaking your complaint? It is easier to detect a violation of a product claim than a method claim.
The EPO is developing a consistent and predictable approach to examining AI-related inventions. But the EPO’s pharmaceutical practice is even more established!
You can claim a product without having to disclose the AI ââ/ ML method used in its development.
A key question for innovators in this field is whether they will need wet lab data to patent a pharmaceutical product in Europe. The sufficiency requirement of the claims for both pharmaceutical compositions and second medical use claims can be summarized as follows from T 1616/09:
For a pharmaceutical composition claim, the request must:
Allow a person skilled in the art to produce the composition and
Leave no doubt as to its use in therapy.
For a second medical use claim, the request must:
Disclose composition in an enabling manner and
plausibly disclose the suitability of the composition for the claimed treatment.
An application without wet lab data relying solely on AI-based predictions will have a difficult course to grant and may attract opposition after grant. If possible, create your new chemical entity or obtain clinical data to support your proposed reused drug. Then you can omit the AI ââ/ ML tool from your patent application.
The EPO’s approach to AI inventions: consistent and predictable
The EPO Guidelines make it clear that AI / ML is currently treated by the EPO as a mathematical method and that mathematical techniques such as AI / ML can contribute to technical character. There is a two-hurdle approach, described below, which gives applicants the tools they need to make a decision to file in Europe that they can have reasonable confidence in.
First of all: does the claimed object have a technical character as a whole? For that, you simply need “technical means”; in other words: a computer. Make sure that your complaint eliminates this obstacle by explicitly stating that the methods are “computer implemented” or by clearly defining a processor, server, device, network, and so on.
Second: does the mathematical method itself contribute to the technical character? Two alternative arguments have been developed in EPO case law that applicants can rely on to ensure that a mathematical method, for example an AI method, contributes to the technical character of an invention. The most relevant argument for AI methods in the MedTech sector concerns the technical application of the mathematical method. In summary, the EPO considers that an IA / ML or a mathematical method can contribute to the technical character of an invention due to its application to a specific technological field. During the prosecution, you will need to be able to argue that the method, in the context of the invention, serves a technical purpose. This technical purpose must be sufficiently specific. Examples from case law and review guidelines include enhancing or analyzing digital images (eg segmentation of medical images); providing medical diagnosis by an automated system processing physiological measurements; using a neural network in a cardiac monitoring device to identify an irregular heartbeat; and providing an estimate of the genotype based on analysis of samples DNA. In addition, the claim must be “functionally limited” to the technical objective. In other words, it is not enough to simply define a technical objective in the preamble of your application. Instead, the mathematical method should be sufficiently linked to the technical purpose, for example by ensuring that the claim defines the data processing that is relevant to the technical purpose, and that the resulting result clearly contributes to the technical purpose.
Of course, while applications that can overcome these obstacles should not expect EPO eligibility issues, this does not necessarily mean that applications that can not clear these obstacles do not stand a chance at the EPO. Contact us and we can give you a go / no-go assessment before filing in Europe. We love challenges!
Stronger, faster and cheaper patents in Europe
What action can you take now? You could:
Review and update your advice to US external lawyers regarding information to be included in AI requests when writing (e.g. technical application / purpose, types of claims)
Review and update invention entry forms for capture information on technical effects and everything is possible technical purposes
To consider co-drafting with a European lawyer to ensure that the correct language is included in the project
To consider pre-filing reviews and modifications before filing in Europe
Consider a go / no go evaluation to allow you to focus your resources on the cases that are most likely to be successful in Europe.