PatentGem and Dexterous Catheters receive an Ignition grant

© 2023 EPFL

© 2023 EPFL

The EPFL Startup Launchpad has granted two Ignition grants of CHF30,000 each to EPFL-based startup projects PatentGem and Dexterous Catheters.

PatentGem

The EPFL Startup Launchpad is pleased to announce they have awarded a CHF30,000 Ignition grant to PatentGem - an EPFL-based technology that will offer companies a radically new method of detecting potential patent infringements.

Most large companies own thousands of different patents. These range from patents on specific products, to processes or methods, to algorithms, or even design and user experience. These patents are strategic assets, helping them maintain a competitive edge, generate revenue, and protect their investments in research and development. In order to safeguard these patents, many companies use specialised software to monitor new products or published patent applications to keep an eye on potential infringements. While it’s hard to quantify the cost of IP infringement, it can easily run into the hundreds of millions – including lost revenue, legal costs, and resources spent on litigation.

The largest challenge these companies face is that current software for recognising IP infringement are good at searching the patent and scientific literature, but they fail to consider the wealth of information available outside of patent databases. This means IP software may fail to detect third-party infringements or may even expose their companies to the risk of infringing other patents.

The team at PatentGem has combined AI and large-language models to develop a technology that will support companies in detecting potential patent infringements. The technology is built on models trained in-house and is independent of other external large-language models (such as GPT-4).

The team at PatentGem will use the grant to improve their software while exploring other funding avenues. They will then work towards incorporation and their first pilot projects.

Contact:

Prakhar Gupta, Scientist at the Chair of Science Technology and Innovation Policy (STIP)

Dexterous Catheters

The EPFL Startup Launchpad has awarded a CHF30,000 Ignition grant to Dexterous Catheters - an EPFL-based spin-off project from the Laboratory of Intelligent Systems (LIS) that looks to improve the success rate and reduce surgery time of cardiac ablation procedures by introducing novel smart catheters.

Abnormal heartbeat, also known as arrhythmia, affects 38 million people and is responsible for around 300,000 deaths yearly. 360.000 cases of arrhythmia are treated every year by cardiac ablation. During such a procedure, a minimally invasive device called a catheter is guided inside the heart to scar the human tissue and normalize the heartbeat. Nowadays, cardiac ablation procedures are performed more frequently with robotic systems that guide catheters using magnetic fields since this approach, also called remote magnetic navigation, is safe for doctors and enables precise control of a catheter. However, these procedures still have a relatively low success rate (13-40%) and patients often undergo several surgeries. This low success rate is in part due to the design of existing catheters. While they are soft enough to navigate the complex structure of the heart, they often buckle under pressure and can only bend in one way – making it a challenge to position the catheter and effectively scar the heart tissue.

Dexterous Catheters uses a technology that’s been in development at EPFL since 2019. It allows the catheter to become rigid or soft on command. These smart catheters can be navigated precisely when soft, and once in place and made rigid, can apply up to 5x more force to the human tissue resulting in fast and high-quality scar formation.

The team at Dexterous will use the grant to validate their value proposition and test the catheters in vitro with leading surgeons. Once the technology has been validated, further funding will allow the development of the final device for preclinical trials.

Contact:

Yegor Piskarev, Doctoral Assistant at EPFL LIS