應用於神經訊號感測之可撓式腦皮質電極與微探針陣列整合、設計與驗證

Abstract

本論文發表一應用於神經訊號感測之可撓式腦皮質電極與微探針陣列整合與設計，達到可同時量測大腦皮質層表面與皮質層內之生理電訊號 。此感測電極主要利用微機電技術製造，考量生物相容性、可撓性與長期植入對大腦的傷害，選擇聚醯亞胺做為大腦皮質波電極與微探針電極陣列之基底，電極製作完成後利用後端封裝技術，將其封裝並整合大腦皮質波電極與微探針電極陣列，其阻抗量測值分別落在4K-6KΩ與118K-168KΩ，符合大腦皮質表面與局部場電位的量測需求，並於實際鼠腦中成功量測4通道之大腦皮質波訊號與14通道之局部場電位。 由於聚醯亞胺基底之微探針電極的彎曲問題，實際植入量測鼠腦時難以將探針電極穿透置放於特定皮質層內分層位置，本論文另外設計與製造一矽基板式微探針電極，其厚度僅26 µm，大幅改善傳統矽基板式電極厚度太厚的問題，且改善電極的因應力不對稱所造成的彎曲問題。並於實際植入鼠腦大腦皮質層聽覺區的第三與第四層後，利用不同聽覺頻率10 kHz與16kHz，以及不同強度0dB與-40dB刺激，記錄到在大腦皮質層聽覺區內的局部場電位反應。最後利用獨立成分分析，成功地投影還原出不同的聽覺誘發電位的訊號源。

This work presents a monolithic neural sensor integrated with an electrocorticographic (ECoG) electrode and micro-probe array. The polyimide (PI)-based ECoG electrode and micro-probe array were fabricated using a customized Micro Electro Mechanical Systems PI process. Proposed sensor can record 4-channel ECoG signals and 14-channel local-field potential signals simultaneously. The selected materials of the fabricated sensor are biocompatible and flexible in order to minimize the damage in long-term implantation. The structure strength has been proof in agar and rat brain for efficient penetration. The impedance of the ECoG electrodes and micro-probes range at 4k-6kΩ and 118k-168kΩ, respectively, and are both fully characterized for neural signal recording. Successful in vivo recordings demonstrated the feasibility of proposed sensor in an awake rat. Due to bending problem of the PI-based micro-probe array, it is hard to penetrate into the desired position of the rat brain. This work also presents a new silicon (Si)-based probe, which greatly reduces the thickness of traditional Si-based probe to 26 µm and possesses enough rigidity for tissue penetration. Successful in vivo recordings for LPF in an awake rat. In the auditory stimulus experiments, the proposed Si-based probe penetrated into the primary auditory cortex of rat brain. Under different frequency (10 kHz and 16 kHz) and different strength (-40dB and 0 dB) of auditory stimulus, the probe can successfully record the auditory response from each channels. By using independent component analysis, the sources of cortex response to auditory stimulus were also successfully derived from the independent components by using back project method.