AI is the next big thing, and most developers want to protect those programs and systems. The answer is intellectual property protection and specifically patent protection.
The post AI, IP, And Your Healthcare: And You Thought The Doctor’s Exam Was Intrusive! appeared first on Above the Law.
Imagine pulling up to a new restaurant on Friday night. It has had wonderful reviews, and you are excited to try it. But a minute after you are seated, you receive an urgent text from the artificial intelligence (AI) app on your phone telling you what you can order. The app has analyzed your medical records, it knows where you are from the phone’s GPS, and it has already evaluated the menu and the nutritional profiles for each appetizer and entree. Wait, it now tells you to simply go home because it has also evaluated the prices and your credit card and bank balances!
Did that describe a dystopian future in which we have lost individual control, or a utopian future in which we are given sophisticated analysis in an instant? What if your AI app also kept a score on how often you complied with its suggestions and sent that score to your health insurance company?
AI and Medicine 3.0
AI is having a substantial impact on healthcare. Numerous AI products have been developed and deployed to assist with the reading of X-rays, CAT scans and pathology studies. In every case, the program is a tool to assist a physician make a final decision. And as these AIs continue to improve, they still are only part of Medicine 2.0 – a term coined to mean the traditional diagnosis and treatment of disease after the patient has gotten sick.
Medicine 3.0 refers to a practice that reduces the risk of a patient developing a disease or disorder. It will look more deeply at genetic information and lifestyle modification for each patient to produce a more focused plan to improve “health span” as well as lifespan. AI, such as the AI example above, should be the perfect tool for evaluating the large data set that represents each of us.
What motivates the development of such tools? We might think it is driven by a universal concern for our fellow man. But, in some cases, it is driven by greed and the wealth that the developer hopes to acquire. AI is the next big thing, and most developers want to protect those programs and systems. The answer is intellectual property protection and specifically patent protection.
Patents and healthcare AI
A patent is a grant from a government to an inventor that provides him with the right to stop others from making, using or selling the patented invention. In the United States, the Patent and Trademark Office (PTO) is responsible for evaluating whether an invention is patentable and then issuing a patent for those that pass its examination. The patent is effective for a period of twenty years from the date the application is filed.
An invention must clear three hurdles to pass the PTO’s examination. First, the invention must be novel – meaning that no other disclosure predates the patent application’s filing date that describes the same invention. Second, the invention must be “nonobvious.” Even if the invention is novel, it must display some aspect that is inventive and cannot be an obvious variation over an existing invention. Finally, the invention must fall into a category of “patentable subject matter.” The patent laws define these categories as machines, processes, articles of manufacture and compositions of matter. An AI invention is subjected to the exact same requirements as any other invention presented to the PTO. Of those three requirements, the most difficult is establishing the AI as patentable subject matter.
Patentable subject matter (PSM)
It seems like in healthcare AI is a “process” – that is a series of steps that takes you from an input to an output. But how often have you heard that machine learning creates its own connections and that the developer does not necessarily know how the AI determines its output. What kind of process is involved if the developer can’t even describe it?
Another troubling attribute for AI, and computer software in general, is that it seems to simply replicate how a human would think through the same problem. Is it really an invention if it only does the thinking for us at a much faster pace?
The PTO is left the job of determining when an AI invention is PSM and when it falls short.
There is a two-part PSM test used by the PTO. The first question is whether the claim made by the inventor is only for an abstract concept. An abstract claim might be for ownership of any AI that improves health. This is simply too vague to obtain protection. Assuming that the inventor claims more, the second question is whether there is “significantly more” added. So, the AI example might qualify as PSM if it was clear that the AI evaluated a location, identified a restaurant from that location, downloaded a menu from that restaurant’s website and analyzed the nutritional content of each offering against a nutrition database. Then, it compared those offerings against an ideal diet for the user based on his last CBC and an evaluation of any medications being taken for contraindications. Each offering would obtain a score and only those above a certain score could be recommended. If none could be recommended, then an alternative plan would be developed (perhaps with AI).
Finding the right balance
Every patent is for a unique idea. That requires careful consideration of how to describe and claim the invention. Too much detail – so you have “something more” – may make the claim so narrow that it has little economic value. A very detailed claim is easier for a competitor to sidestep. In contrast, a more abstract claim may not include enough detail to satisfy the “something more” requirement.
The good news is that the PTO is finding its own right balance in how it examines these incredibly important patent applications in the AI/Healthcare field. According to my recent analysis of the PTO database, approximately 50% of patent applications in this field are issuing as patents. Also, the number of applications are beginning to increase. Almost 9000 patent applications were filed in 2022 related to AI in healthcare.
Note that the rate of issuance appears to be in decline. But this is more likely an indication that some cases from 2022 and 2023 are still being examined. Likewise, the number of applications shown for 2024 only reflects the number filed by mid-year.
Conclusions
The economic value of a great patent can be enormous. The ability to charge customers more for a AI provided service improves when you have a patent(s) to prevent your competitors from introducing the same service. Likewise, understanding the patent positions of a competitor can help your team move its own projects forward more efficiently. However, the path to get a patent is difficult and requires a sophisticated understanding of the law around PSM.
AI is changing healthcare today. There is no stopping its influence and as the systems become more robust, we will all benefit from improved Medicine 3.0.
Photo: tadamichi, Getty Images
David Carstens
David Carstens distinguishes himself not only through his comprehensive knowledge of legal protection of Intellectual Property (IP) but also through his innovative approaches to IP strategy and valuation. With an educational foundation that is as diverse as it is solid — holding bachelor’s degrees in both Electrical and Mechanical Engineering from the University of Texas at Dallas and Texas A&M University, respectively, a J.D. and an MBA from Southern Methodist University, along with completing the General Management Program at the Wharton School at the University of Pennsylvania — David offers a distinctively strategic perspective in this specialized legal domain.
His multifaceted expertise demonstrates his capacity to transcend traditional legal strategies, offering his clients not just defense, but a competitive advantage in various industries including technology, medical devices, cosmetics, and telecommunications. David is a founding partner of Carstens, Allen & Gourley, LLP and has been a pivotal figure on multiple boards. His ability to navigate the complexities of IP law, combined with his technical and business acumen, places him at the forefront of the field.
This post appears through the MedCity Influencers program. Anyone can publish their perspective on business and innovation in healthcare on MedCity News through MedCity Influencers. Click here to find out how.
