擴散式近紅外光斷層掃描系統是新穎的醫學造影技術，此掃描系統的優點是它的安全性高，非侵入式的偵測特性。由於特定波段NIR 光源對於病變組織，因吸收及散射係數差異，而具有效的診斷能力，使其相較其它斷層掃描系統更能有效發現病變組織所在。本論文主要探討在一個高解析度的旋轉機構對人造假體做一個標的物和兩個標物的分析。針對標的物在不同因素下(包括標的物對假體中心偏移量，標的物和假體背景濃度對比度，標的物和假體相對的尺寸比)所造成的影響去做討論。分析所得結論，可用以預測使用此一斷層掃描系統在真實組織下進行量測時，可能的診斷極限，也可提供醫生使用此系統在醫療診斷上之瞭解。 研究中，當NIR 光源位置在絕對座標0 度時，對一個標的物的非均質人造假體進行探討。分析結果可以發現 ⣶98; 當標的物濃度大時，對標的物位置偏移變化時可以有效的辨別出;但對標的 物和實體相對的尺寸做變化時，不易辨別出此變化。 ⣶98; 當標的物位置在不深的情況下，對標的物濃度做變化時和標的物尺寸做變化 時，可以辨別出其中的變化。 ⣶98; 當標的物的尺寸大時，對標的物濃度做變化時，可以辨別出此變化;但對標 的物位置偏移的變化不容易辨別出。 此外當NIR 光源位置在絕對座標0 度時，對兩個標的物 (I0 和I4)的非均質人造假體進行探討。分析結果可以發現 ⣶98; 當標的物0 (I0)的濃度大時，對標的物0 (I0)的位置偏移變化時可以辨別出。 ⣶98; 當標的物0 (I0)位置在不深的情況下，對標的物0 (I0)的濃度變化時可以辨別出，不過是當標的物4 (I4)的濃度小的情況下才成立。 綜合一個標的物和兩個標的物的非均質人造假體分析結果，可以發現當NIR光源位置在0 度和180 度時，對標的物偏移量和標的物和實體濃度對比度的分析結果很類似。本文也利用指數形式的方程式模擬光在介質中的能量分佈，並和實驗人造假體所得的能量分佈做比較討論。 Near infrared (NIR) diffuse optical tomography (DOT) is a new medical imaging modality, the advantages are safe (longer wavelength) and non-radioactive, in spite of its low spatial resolution at the current phase. Due to the diagnostic potential of NIR,the abnormality of tissues in NIR light is to allow earlier detection rather than most other imaging modalities. This thesis describes the NIR Electro-optical measurement for Intralipid phantoms using our own developed NIR DOT scanning instrument with highly spatially angular resolution, in order to investigate the influence of different factors such as the off-boundary, the volume density (v.d.) of inclusion, and the size ratio. The influence of each condition also can reveal the real condition of biomedical tissue, and this information can offer the doctor for diagnosis. This thesis discusses Intralipid phantoms which are the homogeneous background with one inclusion or two inclusions. For the homogeneous background with one inclusion and the source position at s0 , the contrast is large, the resolution of inclusion position is good but the resolution of inclusion size is bad. The off-boundary is small, the resolution of contrast and inclusion size are both good. The inclusion size is large, the resolution of contrast is good but the resolution of off-boundary is bad. For the homogeneous background with two inclusions and the source position at s0 , the contrast (I0) is large; the resolution of inclusion position is good. The off-boundary (I0) is small confined to the non-directly detected inclusion (I4) with small v.d.; the resolution of contrast is good. The common conclusion for the homogeneous background with one inclusion and two inclusions can be obtained. Those phenomena are similar in some condition, especially for source position at s0 and Sπ. This thesis also simulates a power curve by using an exponential form. The light trajectory of model in this thesis was set as straight line, unlike the real trajectory. The outcome does not seem well. From this thesis, it can give a general concept about the influence of different factors. Furthermore, the original experimental data can be used as the input and the calibration of image reconstruction.
