近紅外光擴散光學斷層掃描(near-infrared diffuse optical tomography, NIR-DOT)系統為一新穎之醫學造影方式，其技術主要是利用近紅外光光源透照組織，量測經組織不同光學特性影響而產生之光訊息變化，進一步經計算獲得照射斷面光學係數(散射係數、吸收係數)分佈之成像，反應出內部構造之組織光學特性，其原理歸因於不同部位之生理組織具結構與組成成份之差異，使該光學特性有所不同。 本研究採頻域量測機制獲得光強度與相位差資訊，以此反算待測物之光學性質，並藉由影像之重建計算獲得待測物之內部結構資訊，進而判定待測物內部是否含有異質物(病灶)。首先，以豬肉之瘦肉與脂肪組織製作仿體，藉此模擬乳房光學特性﹔且為製作不同病灶狀況之乳房結構，以矽膠、二氧化鈦粉末與墨水為仿體材料，參考女性乳房之外型結構製作固態乳房仿體。 由實驗結果可驗證量測系統的檢測特性與預期情況相符，且對於不同光學特性之真實組織的檢測，能有效判別出其差異性，此外藉由設計不同型態乳房仿體實驗，比較量測系統對於仿體內不同異質物(腫瘤)對比度、大小及位置的檢測能力，且重建光學係數分佈之影像，由一維剖面曲線觀察重建影像對異質物的檢測結果，可得知乳房仿體中異質物位置對於量測特徵量測的影響最巨。 Near-infrared diffuse optical tomography (NIR-DOT) is a new non-invasive and non-radiation biomedical imaging technique. This NIR-DOT technique uses the near-infrared light illuminating the tissue and measures the transmitted light from the tissue boundary. Using the reconstruction algorithm with the measured data, one can estimate the optical coefficients (scattering coefficients, absorption coefficients) distribution inside the tissue. This is because the optical coefficients vary with the structure and composition within the tissue. In this study, we use frequency-domain scanning system to measure the intensity and phase difference of light propagating through tissue-like phantom, and determine whether heterogeneous inclusion (lesion) appears in phantom or not by images reconstructed from measurement data. In experiment, the tissue-like phantom was made by different type of pork to simulate the optical properties of breast. In order to construct breast-like phantom with different condition of lesion, the solid breast-like phantom was also fabricated by silicone, ink, and titanium dioxide powder, according to the structure of female breast. System verification is first performed by comparing the measurement data from homogenous phantom and the predicted data from the theory model, and judging from the measurement with tissue-like phantom shows that the measurement system is able to differentiate the optical properties in tissue. Moreover, the detect-ability of measuring system is confirmed by comparing the measured data with breast-like phantom under conditions of different inclusion contrast, size, and location. Observation from 1-D optical property profile of images reconstructed from such measurement data shows that the inclusion location is more dominant factor affecting the measurement result.
