Description du projet
Voir les neuroprothèses sous un nouveau jour
Les neuroprothèses recourent à des électrodes pour communiquer avec le système nerveux. Ces dispositifs peuvent supplanter ou compléter les entrées et sorties du système nerveux afin de restaurer certaines fonctions, comme la vue pour les personnes aveugles. Le projet NeuraViPeR, financé par l’UE, construira une neuroprothèse avec des milliers d’électrodes et développera des algorithmes adaptatifs d’apprentissage automatique pour une nouvelle technologie d’interface cerveau–ordinateur qui sera utilisée pour stimuler le cortex visuel. Les travaux du projet comprennent la création de fines (~10 µm d’épaisseur, < 50 µm de largeur) sondes flexibles n’entraînant que des dommages minimes au tissu et la conception de nouveaux revêtements d’électrodes stables malgré la stimulation électrique répétée à long terme. Les algorithmes d’apprentissage profond développés transformeront les séquences vidéo en modèles de stimulation pour le cortex.
Objectif
Approaches that aim to restore vision for blind individuals with electrical stimulation of the brain have hit a technology wall. Existing systems only stimulate a small set of neurons in the brain, and interfaces have a longevity of only a few months. NeuraViPeR aims to lay ground-breaking foundation for a radically new paradigm which consists not only of constructing a neuroprosthesis with thousands of electrodes but also the creation of adaptive machine learning algorithms for a new brain-computer interfacing technology, which will remain safe and effective for decades. Several technological breakthroughs will be established. First, innovative approaches for stimulation with high-electrode-count interfacing with the visual cortex; creating thin (~10 µm thick, < 50 µm wide) flexible probes that cause minimal tissue damage; new electrode coatings that will be stable in spite of long-term repeated electrical stimulation; and novel microchip methods for combining online channeling of the stimulation currents to many thousands of electrodes, combined with monitoring of neuronal activity in higher cortical areas. Second, new deep learning algorithms that transform the camera footage into stimulation patterns for the cortex and that use feedback on recorded brain states and eye tracking to improve perception in a closed-loop approach. The algorithms will extract maximally relevant information to enable blind individuals to recognize objects and facial expressions and navigate through unfamiliar environments. The software algorithms will be translated onto low-latency, power-efficient neuromorphic deep learning hardware, to create a neuroprosthesis system that is lightweight, robust, and portable. NeuraViPeR will tackle these challenges through interdisciplinary teams with complementary expertise in computational, systems and clinical neuroscience, materials engineering, microsystems design, and deep learning.
Champ scientifique
- natural sciencesbiological sciencesneurobiology
- natural sciencescomputer and information sciencessoftware
- engineering and technologyelectrical engineering, electronic engineering, information engineeringelectronic engineeringsensorsoptical sensors
- natural sciencescomputer and information sciencesartificial intelligencemachine learningdeep learning
- medical and health sciencesmedical biotechnologyimplants
Mots‑clés
Programme(s)
Régime de financement
RIA - Research and Innovation actionCoordinateur
8006 Zurich
Suisse