Neuro-ingénierie | EPFL Graph Search

La neuro-ingénierie regroupe l'ensemble des pratiques, disciplines et des technologies visant à améliorer ou modifier les performances cérébrales. Applications médicales La neuro-ingénierie clinique réunit des neuroscientifiques, des docteurs en psychologie cognitive, des informaticiens et des spécialistes du génie des matériaux afin de relever les défis associés à la création d’interfaces entre les neurones et des substrats artificiels afin de restaurer la fonction du système nerveux lésé. Approches exploratoires vignette|100 px|droite|Le caisson d'isolation sensorielle est utilisé pour stimuler l'activité cérébrale. Des chercheurs visent l'augmentation du potentiel humain au moyen de la neuroingénierie. Les techniques les plus fréquemment utilisées sont le neurofeedback, l'isolation sensorielle, l'électrostimulation transcutanée et l'entraînement photique. Dans la culture populaire Dans la culture populaire, le néologisme anglais neurohacking dési

Soft materials and technologies for neural interfaces

Michael Shur

Implantable neural interfaces are an emerging concept which is revolutionizing various domains of medicine and rehabilitation. However, the long-term efficiency and reliability of these devices is often limited, mainly attributed to the fundamental difference between man-made materials and the living neural tissue.The mechanical mismatch between stiff and static materials, such as metals and ceramics, traditionally used in micro-electronics, and the dynamic and soft neural tissue results in various undesirable effects such as performance degradation and chronic inflammation at the device-tissue interface. Therefore, soft neural interfaces are currently an emerging field of research, aimed to overcome these limitations. Two main classes of soft and elastic materials, hydrogels and elastomers, are used to fabricate or modify neural interfaces, envisioning soft and compliant devices, with a minimal mechanical footprint, which can seamlessly interface with the living tissue.In the first part of the thesis, a concept of an ultra-soft hydrogel-based neural interface was developed. It was used to fabricate intra-cortical probes and to study their re-swelling behavior following drying and insertion into the brain. A substantial difference between the initial and re-swollen properties of the hydrogel implants was identified and attributed to the limited amount of free water molecules present in the tissue at the vicinity of the implant. A noninvasive method to track the re-swelling process inside living animal models was developed by using micro-CT imaging and hydrogel implants with an intrinsic X-ray contrast. A preliminary histological examination revealed some undesirable effects of the hydrogelâs re-swelling on the surrounding neural cells, presumably due to hyperosmotic and mechanical stresses caused by the re-swelling process. Influences of geometrical and implantation parameters of the hydrogel probes were also studied, both experimentally and through simulations, providing guidelines for design considerations of future ultra-soft hydrogel-based penetrating probes. The second part of the thesis is dedicated to application-driven design of materials for soft neural interfaces. A conducting hydrogel electrode coating (CHC) was developed and integrated into soft and stretchable microelectrode arrays using screen-printing. Mechanical, electrochemical and biological evaluation of the soft electrode coating revealed its advantages in terms of favorable interaction with neural cells without compromising functional performance. The coating was further validated in two acute recording experiments, in which high-amplitude signals with a good signal-to-noise ratio were acquired. It can be further used to realize entirely-soft neural interfaces for various applications using a wafer-scale high-yield approach.The final part of the thesis focuses on the emerging field of 3D extrusion printing for the fabrication of soft neural interfaces. Several concepts of entirely printable neural interfaces are presented, all utilizing hydrogels and elastomers. A common feature for all the examples is the necessity to formulate printable materials, achieved through the tuning of their rheological properties. Development of several printable hydrogel and elastomer formulations and theirtranslation into functional printed devices was successfully achieved, providing a platform for various in vitro and in vivo applications.

EPFL2021

A lightweight learning-based decoding algorithm for intraneural vagus nerve activity classification in pigs

Silvestro Micera, Leonardo Pollina, Ivo Strauss

Objective. Bioelectronic medicine is an emerging field that aims at developing closed-loop neuromodulation protocols for the autonomic nervous system (ANS) to treat a wide range of disorders. When designing a closed-loop protocol for real time modulation of the ANS, the computational execution time and the memory and power demands of the decoding step are important factors to consider. In the context of cardiovascular and respiratory diseases, these requirements may partially explain why closed-loop clinical neuromodulation protocols that adapt stimulation parameters on patient's clinical characteristics are currently missing. Approach. Here, we developed a lightweight learning-based decoder for the classification of cardiovascular and respiratory functional challenges from neural signals acquired through intraneural electrodes implanted in the cervical vagus nerve (VN) of five anaesthetized pigs. Our algorithm is based on signal temporal windowing, nine handcrafted features, and random forest (RF) model for classification. Temporal windowing ranging from 50 ms to 1 s, compatible in duration with cardio-respiratory dynamics, was applied to the data in order to mimic a pseudo real-time scenario. Main results. We were able to achieve high balanced accuracy (BA) values over the whole range of temporal windowing duration. We identified 500 ms as the optimal temporal windowing duration for both BA values and computational execution time processing, achieving more than 86% for BA and a computational execution time of only similar to 6.8 ms. Our algorithm outperformed in terms of BA and computational execution time a state of the art decoding algorithm tested on the same dataset (Vallone et al 2021 J. Neural Eng. 18 0460a2). We found that RF outperformed other machine learning models such as support vector machines, K-nearest neighbors, and multi-layer perceptrons. Significance. Our approach could represent an important step towards the implementation of a closed-loop neuromodulation protocol relying on a single intraneural interface able to perform real-time decoding tasks and selective modulation of the VN.

IOP Publishing Ltd2022