HearLight Project: A Retrospective

By Olivier Schaetzle, Project Manager - GO 

HearLight was a 3 M€ European Project funded under the H2020 programme of the European Commission. More specifically, it belongs to the FETopen call, FET standing for Future Emerging Technology. The objective of this call was to fund breakthrough ideas at a very early stage and guide them towards the emergence of marketable technology. FETopen calls are now under the European Innovation Council (EIC) banner and are called Pathfinder. 

But let's get back to HearLight. 

First of all, the project was coordinated by Brice Bathellier from the Hearing Institute at Institut Pasteur and gathered 5 more partners: the University of Basel (CH), the Norwegian University of Science and Technology (NO), the Ecole des Mines de Saint-Etienne (FR), NOVAGAN SARL (CH) and the University of Strathclyde (UK). 

And what is it about you might ask? 

You probably heard about cochlear implants, right? These are medical devices currently equipping more than one million of hearing-impaired patients worldwide. And while these implants are bringing a lot of relief to these patients, they have some limitations: not all patients can benefit from them (e.g. patients with inner hear malformation or with damaged auditory nerves) and the quality of perception rapidly decreases in noise or for complex sounds like music 

The HearLight consortium proposed to act directly on the auditory cortex. For this, they used optogenetics, that is illuminating modified neurons with LEDs to simulate artificial sound perceptions in behaving mice. 

While that sounds easy on paper, it means that the consortium had to develop ultrathin, flexible, biocompatible LED displays, that could be placed on the convoluted surface of human auditory cortex to activate precise and rich perceptions. The other side of the project is that they had to decipher the way the brain decodes these stimuli. 

Together, they demonstrated the feasibility of this approach, showing that the cortical implant generates a good quality perception that could be beyond the quality of a cochlear implant. They have developed an encoding model based on a deep learning network model.  

The findings of this project have been patented, and Brice Bathellier is actively working on bringing this technology closer to the market. 

This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 964568.