The large volume and complexity of data available to healthcare professionals is often a goldmine of information for someone looking to correctly diagnose or treat a patient. However, this same data is often cumbersome to work with and can obscure the patterns that may identify a disease or a preferred treatment. Artificial intelligence (AI) and machine learning have the power to peer through the mass of data and pick apart the patterns leading to more accurate and powerful diagnoses and treatments.
The influence of AI and machine learning in medical devices has been growing, as shown by the increasing number of FDA approved AI and/or machine learning-enabled medical devices. The most recently published list of medical devices approved by the FDA stands at ~350 registrations. This list covers medical devices approved between 1997 and 2021. However, there were less than 30 approvals between 1997 and 2015 and approximately 100 individual approvals in 2020. This suggests that the influence of AI and machine learning in the medical device field is accelerating.
The increasing interest in AI/machine learning enabled medical devices raises important considerations for anyone interested in patent protection for such an invention. Initially it seems as though an invention characterised by artificial intelligence or machine learning is excluded from patent protection, as programs for a computer, the presentation of information and methods of diagnosis practiced on the human or animal body are all excluded from patent protection.
Is it possible to patent medical devices?
The short answer is yes, definitely. A medical device that relies on AI or machine learning may have multiple features that can be protected by patents. If the device is an apparatus or piece of equipment, then the device itself can be protected in a patent that describes the device and its inventive features. It is not possible to protect the actual method of using the device on a patient in Europe, however this is possible in the US and Australia (other jurisdictions vary on this point).
Is it possible to patent the AI or machine learning algorithms that underpin many modern medical devices?
Yes. However, the patent has to be carefully drafted, taking into consideration the latest case law in this developing field. The algorithm or machine learning must contribute to the “technical character” of the invention. The technical character of an AI or machine learning enabled medical device may be tough to pin down but there are plenty of examples of how this has been done in cases at the European Patent Office (EPO).
For example, a heart monitoring device that uses an analysis of an electrocardiogram (ECG) by a neural network to monitor changes in a patient’s heart function was considered by the EPO to be a patentable invention. Heart monitoring devices and the analysis of an ECG have been known for many years, however, using a specific neural network to analyse ECG readouts to determine heart function was not known. Therefore, the inventive portion of this medical device was found to be the neural network, not the device itself, and was patentable, despite seemingly being an exclusion. Therefore, the key to ensuring a patent covering an AI or machine learning enabled medical device is granted, is the careful drafting of the patent specification to make sure it does not fall within the excluded subject-matter.
Appleyard Lees IP LLP has a number of patent attorneys who specialise in protecting medical devices and medical technology more broadly. Please contact [email protected] or visit www.appleyardlees.com to find out how we can help secure your innovation.