Êtes-vous un étudiant de l'EPFL à la recherche d'un projet de semestre?
Travaillez avec nous sur des projets en science des données et en visualisation, et déployez votre projet sous forme d'application sur Graph Search.
Recording neural activities plays an important role in numerous applications ranging from brain mapping to implementation of brain-machine interfaces (BMI) to recover lost functions or to understand the mechanisms behind the neurological disorders such as essential tremor, Parkinsonâs disease and epilepsy. It also constitutes the first step of a closed-loop therapy system which employs a stimulator and a decision mechanism additionally. Such systems are envisaged to record neural anomalies and then stimulate corresponding tissues to cease such activities. Methods for recording the neural signals have evolved to its current state since decades, and the evolution still goes on. This thesis focuses on how to eliminate all the wired connections for new generation neural recording systems: implantable wireless neural recording systems with a case study on in-vivo epilepsy monitoring. The scope of the thesis can be defined as wireless power transfer, wireless data communication, biocompatible packaging, and compulsory experiments on the way to human trials. First of all, wireless power transmission is performed using 4-coil resonant inductive link topology which exploits the magnetic coupling phenomena. In addition to power transfer, a reliable DC power supply is generated in the implant by means of a half-wave active rectifier and a low drop-out voltage regulator. The operation frequency, 8.5 MHz, has been optimized by taking tissue absorption and bandwidth limitations for data communication into account. Secondly, wireless data communication solutions have been investigated and two different solutions have been implemented for different application scenarios: First solution is to use load modulation scheme, which actually relies on varying the load according to the incoming neural data. However, there is a trade-off between data rate and power transfer efficiency for this solution, which in return leads us to implement the second solution, dedicated transmitter at a higher frequency. Consequently, a transmitter which can work at MICS (402-405 MHz), ISM (433 MHz) and several MedRadio bands has been implemented to transmit neural data to an external base station which includes a discrete receiver. Following the integration of all electronic circuits which have been fabricated using UMC 180 nm MM/RF technology, the implant has been packaged using biocompatible polymers (PDMS, medical grade epoxy, and Parylene-C). Packaging provides bidirectional diffusion barrier feature which enables in-vitro and in-vivo experiments to be conducted. Finally, three levels of experimentation have been conducted to validate the operation of the system: in air for electrical characterization, in a tissue-mimicking solution in-vitro characterization, and in a mouse brain for in-vivo characterization.
Mahsa Shoaran, Uisub Shin, Gregor Rainer, Mohammad Ali Shaeri, Amitabh Yadav