| 摘要: | 乳癌是女性最常發生的癌症之一，且發生率有逐年攀升之趨勢。然而統計報告顯示，前期乳癌(第零到二期) 的存活率高達90%以上，說明了乳癌盡早診斷與治療的重要性。目前現有的臨床檢測設備，如：乳房X光攝影(X-ray mammography)、超音波檢測(Ultrasound) 及磁振造影(Magnetic Resonance Imaging)，有著輻射疑慮、操作者技術、價格昂貴及適用對象限制等問題；因此，擁有非輻射、非侵入性、適用年齡層廣泛等優勢的頻域式近紅外光擴散光學斷層影像(Near Infrared Frequency domain Diffuse Optical Tomography, NIR FDOT)便開始成為熱門的研發對象。
NIR FDOT係以25 - 1000 MHz交流電源驅動700 - 900 nm近紅外光照射待測物(如乳房)，並利用光接收器(如光電倍增管)接收從待測物表面射出的光資訊，即因腫瘤組織對於近紅外光之吸收及散射特性與正常組織的差異而產生的光衰減量(ΔA)與相位延遲量(Δθ)。使用越多頻率驅動光源可得到越多組的光資訊，並可預期提升重建影像的品質，但也因順序式量測而增加了檢測時間。本篇論文基於多頻驅動光源的作法，實驗中以訊號產生器(AFG3102) 輸出29 MHz方波電源驅動830 nm雷射，量測直徑80 mm，高90 mm，"μa" 及 "μs′" 為0.0074 mm-1及0.85 mm-1之圓柱形光學假體，將不同大小及光學係數對比度之橢球型腫瘤假體嵌於與背景假體相同光學係數之柱條內，以插拔式改變不同大小、對比度腫瘤之離心距，共四個案例。因方波為複合弦波的特性，在同一量測時間下能獲得相較單頻驅動光源數倍的光資訊，實驗中收集方波驅動光源前兩諧頻的光資訊，藉多頻擴散光學斷層造影重建出圓柱仿體內部"μa" 、 "μs′" 光學分佈影像，並以CSD量化後進行方波與單頻驅動光源之重建影像比較，預期大顆腫瘤假體的"μa" 、 "μs′" 重建影像之CSD量化值會較小顆高；腫瘤假體在位置淺處的"μa" 、 "μs′" 重建影像之CSD量化值會較深處高；頻率高之單頻驅動光源量測在"μs′" 重建影像之CSD量化值應較低頻優異；方波驅動光源的 "μa" 、 "μs′" 重建影像CSD量化值應較高、低單頻驅動光源優異。而結果可發現，因訊號產生器輸出的方波並非理想形式及實驗設備的限制，使方波驅動光源未有顯著的效果，期待未來能以合成複合週期訊號驅動光源量測進行改善。
;Breast cancer is one of the most common cancers among women, and the incidence rate is increasing year by year. However, the statistical report shows that the survival rate of early breast cancer (stages 0 to 2) is higher than 90%, indicating the importance of early diagnosis and treatment of breast cancer. Existing clinical testing equipment, such as: X-ray mammography, Ultrasound and Magnetic Resonance Imaging, have radiation doubts, highly depending on operator’s skill, expensive price or restrictions on applicable objects. Therefore, frequency-domain near-infrared frequency domain diffuse optical tomography (NIR FDOT), which has the advantages of non-radiation, non-invasiveness, and can be adapted on broad age range, recent year it has become a popular research object.
The NIR FDOT is applying 25 - 1000 MHz AC power to drive 700 - 900 nm near-infrared light to measure the object (such as breasts), and uses a light receiver (such as a photomultiplier tube) to receive the light information comes out from the surface of the object, that is the amount of light attenuation (ΔA) and phase delay (Δθ) due to the difference between the absorption and scattering characteristics of tumor tissue and normal tissue. Using more frequencies to drive the light source can obtain more sets of light information, and it is expected to improve the quality of the reconstructed image, but it also increases the detection time due to sequential measurement. This thesis is based on the method of driving the light source with multiple frequencies, yet in the experiment, the function generator (AFG3102) is used to outputs a 29 MHz square wave to drive a 830 nm laser, measuring 80 mm in diameter, 90 mm in height cylindrical optical phantoms, its "μa" and "μs′ " are 0.0074 mm-1 and 0.85 mm-1, embedding ellipsoid tumor phantoms of different sizes and contrasts of optical coefficients in columns with the same optical coefficients as the background phantom. There are four cases in which the off center distance of tumors of different sizes and contrasts is changed in a plug-in manner. Since the square wave is a characteristic of a composite sine wave, it can obtain several times light information compared to a single-frequency driving light source at the same measurement time. In the experiment, the optical information of the first two harmonic frequencies is collected, and using developed reconstructing software to reconstruct optical distribution images of "μa" and "μs′ " inside the cylindrical phantom, and quantified by CSD value. The reconstructed images of the square wave and the single frequency drive light source were compared. Expected CSD value of large tumor will be higher than small one; and the CSD value of the tumor placed shallow will higher than the one in deep; the CSD value of high frequency measurement should be higher than low frequency measurement; the CSD value of square wave measurement should be higher than both high and low single frequency measurement. As a result, it can be found that the square wave output by the signal generator is not an ideal form, also some restrictions of equipment, so that the square wave measurement has no significant effect. It is expected that the composite periodic signal driving light source measurement can be applied and improved the result in the future. |
