基於電場成像與特定方法以模擬人工電子耳之組織阻抗網

Abstract

具有極重度的聽障患者可藉由人工電子耳獲得益處，而其修復部分的聽力是藉由植入一個電極陣列，以此取代受損的毛細胞，直接刺激聽覺神經纖維。即使人工電子耳是目前最成功的神經裝置，但其在控制內耳蝸之電流的能力上，仍存在些限制，而這可能會導致電刺激性能不佳，因此了解電流的路徑是非常重要的。其次，臨床醫生常無法輕易地識別人工電子耳之電極陣列是否有短路或者異常的電流流向，除了藉由量測電場成像外，但其僅能夠得到粗略的問題所在，而非精準地確認問題。 本篇論文的兩個主要目的是，鑑定一個等效於人工電子耳患者的個人化電阻抗網絡，並使用此網絡確認同一患者在臨床上是否有任何異常的漏電流問題。我們提出了一種新的混合方法，以電的形式體現人工電子耳系統的特徵；首先，我們定義了組織電阻抗模型的結構，其用於模擬人工電子耳受測者之植入電極的特性與生理解剖，並在先前，採用兩種不同的方法，又稱為電場成像最佳化法與雙電極刺激相減法，以估算組織電阻抗模型的參數；然而，這兩種方法都不足以準確地決定組織的阻抗值與單電極刺激時的電流流向，因此在本論文中，我們提出了一種新的方法，又稱為混合方法，其採用上述兩種方法的優點，以改善結果的準確性與穩定性；此外，藉由此混合方法所得到的人工電子耳之組織阻抗模型，我們能夠更好地理解電流的分布，以及知道是否有任何異常的漏電流。最後，基於此方法得到的參數，我們可創建一個針對個人化的有限元素模型，並改善未來人工電子耳之電極陣列的設計，而所有的結果將藉由電場成像的量測作驗證。

People with severe to profound hearing loss can obtain benefit from the cochlear implants (CIs), which restore partial hearing by bypassing the impaired hair cells and directly stimulating the auditory nerve fibers with an implanted electrode array. Even cochlear implants are the most successful neural prostheses today, there remain some limitations in our ability to control the intracochlear current flow, which could result in poor electrical stimulation performance. Therefore, it is vital to know the paths of current flow. Secondly, clinicians often cannot easily identify if there is any CI electrode array short circuit or abnormal current flow except by measuring the electrical field imaging (EFI), which only gives a rough location of the problem without precisely identifying the problem. The dual-purpose of this thesis is to identify a patient specific electrical impedance network equivalent to a CI patient electrically and to use it to identify clinically whether there is any abnormal current leakage problem with the same CI patient. We are proposing a new hybrid method to characterize the cochlear implant system electrically. At first, we define the structure of the tissue electrical impedance model, which is used to model the properties of the implanted electrode array and anatomy of the CI test subject. Previously, the two methods, referred to as Electrical Field Imaging Optimization Method (EFIOM) and Bipolar Subtraction Method (BSM), are used to estimate the parameters (rLongitudinal, rTransversal and ZInterface) of tissue electrical impedance model. However, both of these methods are not precise enough to determine the tissue impedances and current flows accurately during monopolar stimulation. In this thesis, we propose a new method, referred to as Hybrid Method (HM), which takes advantage of the strength of both EFIOM and BSM to improve the accuracy and stability of the solution. Furthermore, we can acquire a better understanding of the current distribution or if there is any abnormal current leakage from the cochlear implant tissue impedance model using the proposed Hybrid Method (HM). Finally, we can create a patient-specific finite element model based on the parameters (rLongitudinal, rTransversal and ZInterface) obtained from this method and can help improve future CI electrode array design. All of the results are validated by the electrical field imaging (EFI) measurements.