The frontier of MedTech: AI-driven medical devices
November 20, 2023
The remarkable surge in the field of generative Artificial Intelligence (AI) has far-reaching implications for Medical Technology (MedTech). Generative AI refers to the application of artificial intelligence to create, or assist in the creation of, designs based on input criteria and constraints. Generative design techniques are being harnessed to revolutionize the creation of medical devices.
Engineers and designers use AI algorithms to remove unnecessary material or otherwise improve a medical device’s design. Just as an engineer draws upon their training and experience to come up with new solutions, generative AI draws upon far larger databases of information and far more complex calculations than any human can compute to optimize a product’s design – thus the name “artificial intelligence.” AI allows engineers to adapt designs to specific patient needs and complex medical requirements, and to accelerate the development process. This helps patients and medical professionals alike.
What does AI in MedTech innovation look like?
Here are two examples of firms that use AI to significantly improve patient care and outcomes.
Meticuly, a medical technology company based in Thailand, uses generative AI to develop personalized bone implants. Meticuly feeds the patients’ CT scans into an AI deep learning algorithm to design the implant. They then 3D print the generated design. Before AI, the bone implants were molded by hand and required doctors to manually adjust the implant’s shape during the operation to perfect the final fit in the patient’s body. Meticuly eliminated that need and reduced operation time by hours, improving outcomes for patients.
NuVasive, a medical device company based in the United States, uses generative design to create 3D printed porous titanium spinal implants. Titanium spinal implants are usually made of solid chunks of metal. By using generative AI and 3D printing technology, NuVasive can make a sponge-like titanium implant with tiny holes all throughout it. This takes advantage of titanium’s biocompatibility while giving the healing spine more holes to grow through and making the implant more flexible to prevent a subsidence (a common complication of spinal fusion surgeries). When using a traditional spinal implant, the rate of subsidence is around 20%. When using an AI-generated porous titanium implant that rate goes down to 1.1%.
How does Generative AI impact a MedTech company’s IP portfolio?
The use of generative AI in designing medical devices raises important questions and challenges regarding intellectual property rights and ownership of the generated designs: can designs that were generated using AI be patented? What is the IP difference between AI-assisted designs, i.e., designed by a human and refined by an AI, and AI-generated designs?
At the Sixth session of the WIPO Conversation “Frontier technologies – AI Inventions” Meticuly co-founder and CTO, Asst. Professor Chedtha Puncreobutr described Meticuly’s IP portfolio and how it was affected by AI. The technology is made up of the combination of the 3D printed titanium material and the generative AI program coming together to create the final product.
Meticuly protects their material with invention patents and protects their generative AI through Technology Trade Secrets. The product family of 3D printed implants is covered using a mix of invention and design patents. Meticuly holds six patents on different aspects of their products and 12 trade secrets covering their generative AI and other manufacturing processes. This is a good example of the IP portfolio of an AI-assisted design.
For AI-generated designs, determining the rightful IP owner can be challenging. Questions surrounding whether the AI system’s creators, the users of the AI, or even the AI itself can claim ownership and protection for these designs are actively being debated.
The question of ownership of AI-generated designs is significantly influenced by the data used to train the AI systems. AI models learn from vast datasets that can encompass a wide range of pre-existing designs and creative works. In some cases, the data used may contain copyrighted or proprietary information, which raises concerns about the potential infringement of existing intellectual property rights. Additionally, when AI systems generate novel designs, questions arise regarding the extent to which these outputs are influenced by the training data and whether they can be considered entirely original.
These questions were recently brought to a global level of discussion by the Device for the Autonomous Bootstrapping of Unified Sentience (DABUS) AI System. The DABUS AI system was created by Dr. Stephen Thaler and patented in US patent publications: 5,659,666; 7,454,388 B2; and 2015/0379394 A1. DABUS gained worldwide attention for being the first AI system to supposedly come up with two inventions on its own (a fractal drink container and a flickering alert light) and the first AI system to have patents filed on its behalf.
Dr. Thaler and his team filed patents for these inventions in Australia, the European Patent Office, New Zealand, South Africa, the United Kingdom, and the United States. This issue was highly debated around the world, and most countries rejected the filings on the basis that an inventor must be a human. Only the South African Companies and Intellectual Property Commission (CIPC) accepted Dr. Thaler’s application on June 24th 2021 and released a notice of issuance for the patent in July 2021. It is not clear yet how this decision will impact the field. It is possible that this decision could be the catalyst for the development of a sui generis (“of its own kind”/specialized) IP system for AI inventions, but it may also have no impact at all.
AI and MedTech – what happens next?
The intersection of generative AI and medical device design brings about complex challenges related to intellectual property rights, especially when considered within the context of fully AI-generated designs. The case of the DABUS AI system, which sought patent recognition as the inventor, has ignited global debate on the rights of AI entities and the future of IP law.
While the effect of AI on the IP ecosystem remains to be seen, its impact on the MedTech space is already clear. The use of this technology is ushering in a new era of innovation in MedTech and offering tremendous potential to personalize medicine and reduce complications from medical procedures. Recognizing this impact, the World Intellectual Property Organization (WIPO) recently launched the Global Initiative on AI for Health initiative with the International Telecommunication Union (ITU) and the World Health Organization (WHO). Our three UN agencies are coming together to support AI in achieving its full potential to promote improved patient outcomes.
Companies like Meticuly and NuVasive are the first pioneers to use of generative AI to improve patient lives, but they will not be the last. Unlocking this new tool gives us the potential to enhance patient care and outcomes around the world.