French Company Pioneers Artificial Intelligence for Brain Targeting

Published:

Key Points

  • RebrAIn pairs artificial intelligence with health data to simplify and standardize targeting for neurological procedures. 
  • Learn how the technology works for surgical interventions and how the company plans to apply it to focused ultrasound treatments. 
RebrAIn logo

RebrAIn is start-up company that has developed software to help neurosurgeons identify operable brain regions in patients with Parkinson’s disease (PD) and essential tremor (ET). Using an artificial intelligence (AI) algorithm and a collaborative registry of health data, RebrAIn simplifies and standardizes procedures with the aim of increasing the number of treated patients worldwide.  

The company was established in 2021 based on a decade of academic research at Bordeaux University Hospital and the National Institute for Research in Digital Science and Technology (Inria) by Emmanuel Cuny, a professor of neurosurgery at Bordeaux University Hospital and doctor in mathematics Nejib Zemzemi. In just three years, the company’s software was granted the CE Mark and US Food and Drug Administration (FDA) clearance for deep brain stimulation (DBS) treatment of both PD and ET.  

Seasoned focused ultrasound executive David Caumartin recently joined the RebrAIn team as chief executive officer. We spoke with Mr. Caumartin about the RebrAIn technology and how the company plans to apply it to focused ultrasound treatments. 

How does RebrAIn’s technology work?  
The software – called Optim MRI – is an automatic system that allows the physician to target areas of the brain for treatments like DBS, radiosurgery, and hopefully one day, focused ultrasound. The difficulty of targeting for neurostimulation is that the brain has no precise marker for where the patient should be treated, and the treatment area must be precise (within one to two millimeters). For the RebrAIn algorithm, we identified 18 landmarks per hemisphere to create a GPS-like positioning system for the brain. The physician uploads an MR brain image to the platform, and then the platform automatically marks the areas to guide the physician on where to aim the neurostimulation.  

The best analogy I can find is if you imagine traveling to New York City for the first time, and you have no smartphone or GPS. You will spend much more time and stress trying to navigate the city. Optim MRI for neurostimulation is like a GPS system for identifying optimal treatment targets; it removes variants that could slow the process and reduces physician stress.  

How does the software know where to aim for specific conditions? What makes it beneficial to physicians? 
Optim MRI amalgamates past real-life treatment data that is associated with positive outcomes to reproduce what the best physicians do manually. It evens the playing field for physicians who might not be an expert in brain targeting for these treatment modalities. It can also lead to shorter treatment times. 

Why is AI integral to the software?  
AI allows the Optim MRI software to build a digital twin of the brain to inform neurological disease treatment decisions. It allows the addition of complex data that the human brain is unlikely to repeat consistently.  

How can physicians access the software? Is it available today?  
Optim MRI is an SAS, or software as a service. It is a cloud-based platform in which the user uploads the MR sequence along with the disease being treated. The algorithm then marks the MR sequence and identifies the treatment zone. The user downloads the marked MR and uses it as the pre-operative 3D dataset that is uploaded to the treatment device. There is no need for another device – it is entirely software-based. 

To date, the technology has been granted CE mark and FDA clearance for use in DBS treatments of PD and ET. It is available commercially in the US and Europe, and it has been used in nearly 500 treatments. The software has also been tested in about 60 sites worldwide.  

Optim MRI is being tested in a multicentric randomized controlled clinical trial to assess its ability to target the areas without micro electro recording in a full asleep procedure for DBS of PD. We are also beginning a pivotal clinical trial for treating ET with radiosurgery, on a Gamma Knife.   

How did you get involved with RebrAIn? 
I started in the medtech industry in 1994 when I joined GE Healthcare. In 2000, I first heard of focused ultrasound because GE was involved with the launch of Insightec at the European Congress of Radiology that year, and I was always fascinated by the use of focused ultrasound. In 2014, I started as CEO for a public company called Theraclion, and we developed an ultrasound-guided focused ultrasound device for breast tumors, thyroid tumors, and varicose veins. After my time at Theraclion, I became involved with the Focused Ultrasound Foundation and then the opportunity arose at RebrAIn.  

How could Optim MRI impact focused ultrasound treatments? 
Just as we are able to target the areas of the brain for ET and PD for a DBS treatment, we believe we can do the same for a focused ultrasound procedure. Targeting is an important topic of interest, and Optim MRI allows both experienced and inexperienced physicians to confidently target an ET treatment based on clinical data not only MRI morphological imaging. Using AI this way could lead to more centers around the world adopting the technology. It also could drastically reduce the time that a patient stays in the MR machine during treatment and increase confidence in all neurosurgeon potentially interested in beginning MR-guided focused ultrasound treatment for ET.  

How soon could it be available for focused ultrasound?  
We are in the process of raising funds for an FDA filing of Optim MRI for lesioning in patients with ET and a pivotal clinical trial that will likely take place in Spain and a select few sites in the US that have vast experience with treating ET with focused ultrasound.  

What’s next for RebrAIn?  
The focus is to grow the company, both in the US market and to extend beyond DBS into minimally invasive options like radiosurgery and focused ultrasound. We believe the best way to increase the number of patients treated for both PD and ET will be with MR-guided focused ultrasound because it is the fastest, least invasive, and the simplest to do for all healthcare systems around the world. We hope to be able to provide a better, safer, faster solution for neurosurgeons to map the appropriate brain targets. We aim to offer all patients suffering from severe movement symptoms of PD and ET an accessible, safe, and effective treatment solution. 

How do you see AI impacting care in the long term?
AI will play a great role in healthcare because of its ability to read and understand vast amounts of data. One of the major areas is patient selection and segmentation: being able to identify the best treatment and the right steps for each individual. RebrAIn is an important step in that direction, and in 10 years, the field will look drastically different.  

See the RebrAIn Website