三維空間重建暨模擬中耳聽小骨的生物機械模型 及其臨床應用

Abstract

在這個研究裡，包含了以有限元素分析法為基礎，探討如何從高解析度電腦斷層來建立人類中耳模型，並以此模型整合結構音場的耦合來探討乳突竇對聲音傳遞的影響，此外，在臨床上的應用上；例如：建立了軟骨及耳膜耦合模型並決定出軟骨耳膜修補手術的最佳化厚度，及應用在研發新一代助聽器。 首先，吾人採用高解析度電腦斷層擷取受測者顳骨的影像，並利用影像重建的方法，建立三維體積模型，並分析耳膜的厚度、聽小骨的長度，最後在利用電腦輔助分析建立CAD (Computer-aided design) 模型，再利用有限元素分析，解算出malleus及stapes的振動大小，並且和國外學者Nishihara, Huber, Gan, and Sun等人的資料做比較，除了在耳膜厚度的量測上，因為受限於高解析度電腦斷層的解析度，但是這個缺點可以在做有限元素分析時做相互校正給調整出來，其它所有結果是互相一致的，吾人所提出的方法和組織切片方法比較起來擁有快速、低成本、非侵襲性等優點。 第二個研究則進行了利用高解析度電腦斷層擷取受測者顳骨的影像，並利用影像重建的方法，建立三維體積模型，包括了外耳道、耳膜、3個小聽骨、中耳腔及乳突腔的有限元素模型建立，不僅可以用來解釋乳突腔對於聲音傳導的影響，藉由因場結構耦合分析，可以瞭解因為乳突腔的連結，對於聲音的壓力在中耳傳遞的影響。 第三個研究包含了tragus 軟骨的參數度量及其研究，吾人決定出了tragus軟骨的密度及其β-damping值的大小，且其值和頻率的大小有關，在低頻時β-damping接近3 10-4 s然而在高頻時β-damping接近5 10-6 s，並且建立了軟骨及耳膜耦合模型，這個軟骨及耳膜耦合模型可以用來決定手術時採用軟骨耳膜修補手術的最佳化研究 第四個研究包含了在不同耳膜破洞大小時，採用軟骨耳膜修補手術時最佳厚度之決定，依據吾人的第二個研究及創立的軟骨及耳膜耦合模型，建立了三種不同耳膜破洞大小的CAD模型(15%, 55% 及85%)，同樣符合破洞大小的軟骨板在連結至這個模型，藉由有限元素分析的過程，最佳化厚度可以被求出，結果顯示：在小破洞（大約15%）使用厚度1mm可以得到良好的結果，然而在在中等大小破洞（大約55%）使用厚度0.2mm可以才得到良好的結果，如果是更大的破洞（大約85%）就必須使用厚度0.1~0.2mm可以才得到良好的結果，總而言之，採用軟骨耳膜修補手術，需依據不同的破洞大小來決定其使用的厚度。 最後的研究是研發新一代的耳膜上助聽器，目的上想解決傳統式助聽器的一些缺點：回饋現象(feedback)、遮蔽效應(occlusion effect)、在Earlens助聽器上增益不穩定•••等現象，並使用發光二極體及相對應的光二極體作為訊號的接收，最佳化設計包含了：磁鐵的大小、線圈的位置及形狀、光電流的大小。這個致動器(actuator)是預計放在耳膜之上，利用發光二極體產生的光作為致動器(actuator)的能量來源，因為光的能量是正的無法區分出正的訊號及負的訊號，所以我們採用半波整流(half wave rectifier)把訊號分成正、負兩週期訊號，讓正訊號藉由藍光發光二極體並藉由相對應的藍色二極體接受；同樣地，讓負訊號藉由綠光發光二極體並藉由相對應的綠色二極體接受，因為線圈是兩個方向性纏繞，如此一來將產生推拉的力量而使致動器作用，所模擬電磁力的最佳化也可作為新一代助聽器的先驅。

In this research, we first demonstrate how to use high-resolution computed tomography to create human middle ear as a systematic and practical approach. Finite element and multi-body dynamic analysis of this model can be used. The three-dimensional model created by finite element method and predicted umbo and stapes displacement are close to the bounds of the experimental curve of Nishihara’s, Huber’s, Gan’s, and Sun’s data across the frequency range of 100 to 8000 Hz. The second work was to created a three-dimensional finite element (FE) model of the human ear with the external ear canal, tympanic membrane, ossicles, suspensory ligaments/muscles, tympanic cavity and mastoid cavity from high-resolution computed tomography images (HRCT). Acoustic-structural coupled FE analysis was performed to study the sound transmission of the human middle ear. The effect of the mastoid cavity on sound transmission was also highlighted. Pressure distributions in the external ear canal and middle ear cavity at different frequencies were demonstrated. The FE model of the human ear was validated by comparing model-predicted ossicular movements at the umbo with published experimental measurements on the human temporal bone. Our results showed that, first, blocking the aditus improves middle ear sound transmission in the 1500- to 2500-Hz range and decreases displacement in frequencies below 1000 Hz when compared with the normal ear. Second, acoustic pressure distributions in the external ear canal and middle ear cavity depend on frequency. At frequencies lower than 1000 Hz, the acoustic pressures were almost uniformly distributed in the external ear canal and middle ear cavity. At high frequencies, higher than 1000 Hz, the pressure distribution varied along the external ear canal and middle ear cavity. Third, after coupling with the mastoid cavity, the pressure differences in the middle ear cavity were larger than those of the closed mastoid cavity. Finally, there was no significant difference in the acoustic pressure measured at different locations in the middle ear cavity at low frequency. As the frequency increases, the pressure difference between the oval window and round window is noted and increased by 5 dB by blocking the aditus. In clinic, this model can be used to develop a cartilage/tympanic membrane-coupled model for cartilage myringoplasty. Optimal thickness of different sizes of cartilage plate was obtained using finite element analysis. Parameters of cartilage were determined by curve fitting and cross-calibration. Our results show the β-damping value of cartilage plate depends on frequency. The value of β damping was close to 3 10-4 s at lower frequencies and 5 10-6 s at higher frequencies. Three different sizes of TM perforation (15%, 55% and 85%, representing small, medium and large perforations respectively) were created in the pars tensa. A cartilage plate was used to repair the eardrum perforation, and the new tympanic membrane-cartilage coupled complex was loaded into our 3-dimensional biomechanical model for analysis. Our results show that firstly, in cases with 85% perforation, the frequency-amplitude responses that were most similar to natural TM at lower frequencies were for graft thicknesses of 0.2 mm, and for 0.1 mm at higher frequencies. Secondly, in cases with 55% posterior perforation of the TM, assessment of the predicted vibration amplitude of different thicknesses of the cartilage plate showed that a cartilage plate of < 0.2 mm had a frequency response function similar to that of a natural TM in umbo and stapes footplate displacement. Finally, for a central perforation involving 15% of the tympanic membrane, a cartilage plate of < 1.0 mm showed a frequency response function similar to that of TM in umbo and stapes-footplate displacement. Based on our biomechanical analysis, the optimal thickness of a cartilage graft for myringoplasty seems to be 0.1 to 0.2 mm for medium and large TM perforations. For small perforations, a cartilage < 1.0 mm is a good compromise between mechanical stability and low acoustic transfer loss. We also design a new method of transducing sound to the human middle ear. The new hearing aid uses light from light emitting diode as the transferring media of input signals. Then the receiving photodiode generate currents and signals to supply the power of actuator. The new type of vibration actuator is composed of wound coil, a permanent magnet, aluminium ring, latex membrane and two photodiodes. The actuator is placed on the tympanic membrane and maintains its position by mineral oil. Finite element analysis is performed to optimize the electromagnetic force.