擴散光學斷層造影(diffuse optical tomography, DOT)，即將近紅外光進入組織中，由光偵測器蒐集光資訊，進行光學係數分佈影像重建。進行影像重建時，分為前向計算與逆向反算兩個部分。前向計算藉擴散方程式呈現光在組織的傳遞情形，獲得不同位置的光資訊；逆向反算則通過牛頓法進行疊代，利用最小化量測資料與前向計算的差值，重建組織光學係數分佈，進行腫瘤大小及位置的判斷。逆向反算中，由於擴散係數遠大於吸收係數，導致吸收係數主導重建圖像，故需加入正規化項，平衡兩光學係數之數量級，使吸收係數不會主導重建；同時逆向反算中具有病態特性，故需加入正則化項穩定重建結果。 為了要方便執行影像重建，故本論文將分別整合實驗室現有的DOT、FDOT影像重建程式並設計使用者介面，方便後人在進行影像重建時使用；也在多波長組合挑選的程式中設計使用者介面，以便依組織組成濃度，從不同波長光源中挑選最佳的波長組合。 1) 軟體改進方面： 在軟體改進方面，分別改善網格中光源與光偵測器位置的判斷；修改原三維多頻率光資訊影像重建程式，找前向計算與量測資料的實數與虛數部分差異時的計算錯誤；增加依仿體輪廓進行一維等比例剖面呈現影像重建結果的方式。 2) 軟體驗證方面： DOT部分以圓柱、半橢球仿乳和不規則形仿體驗證，確認本軟體在使用單頻率與多頻率光資訊進行二維與三維影像重建的能力。FDOT部分使用圓形仿體驗證，確認本軟體影像重建的能力。多波長組合挑選參考女性乳房組織組成濃度，作為驗證本軟體在市售波長挑選多波長光源組合時的依據。 ;Diffuse optical tomography (DOT), in which laser light enters the tissue, and the light information is collected by detector. There are two parts in the image reconstruction: forward calculation and inverse reconstruction. The forward calculation uses the diffusion equation to present the light transmission in the tissue and obtain the light information at different positions; the inverse reconstruction is to iterate through the Newton method, and the distribution of optical coefficients in tissue is reconstructed by minimizing the difference between the measured data and the forward calculation to determine the tumor size and location. However, in the inverse reconstruction, since the diffusion coefficient is much larger than the absorption coefficient, the absorption coefficient dominates the reconstructed image. Therefore, it is necessary to add a normalization term to balance the two optical coefficients to prevents the reconstruction from being dominated by the absorption coefficient; at the same time, due to the ill-conditioned of the inverse reconstruction, a regularization term needs to be added to stabilize the reconstruction result. In order to make image reconstruction more convenient, in this paper we integrate the existing DOT and FDOT image reconstruction programs in the lab and design the user interface to facilitate the image reconstruction by future users; it is also designed and used in the program for multi-wavelength combination selection. The user interface is also designed in the multi-wavelength combination selection program to select the best wavelength combination from different wavelength light sources. 1) Software improvements: The way of determine source and detector position in the mesh was improved; modify the error in the original 3D multi-frequency image reconstruction program when finding real and imaginary parts of the difference between forward calculation and measurement data; add the way to present the image reconstruction results in a one-dimensional equal scale profile according to the contour of the phantom. 2) Software verification: In the DOT section, cylindrical, semi-ellipsoidal breast-liked and irregular-shaped phantoms were used to verify the ability of this program to perform 2D and 3D image reconstruction using single-frequency and multi-frequency light information. In the FDOT section, circular phantoms were used to verify this program image reconstruction ability; In the multi-wavelength combination selection, the concentration of female breast tissue composition was used to verify the program′s ability to select multi-wavelength light source combinations from commercially available wavelengths.
