指導教授：劉子銘臺灣大學：醫學工程學研究所謝佳宏Hsieh, Chia-HungChia-HungHsieh2014-11-302018-06-292014-11-302018-06-292013http://ntur.lib.ntu.edu.tw//handle/246246/264443血液查常規檢查是重要的醫療篩檢項目，透過血液抹片的顯微影像來了解血球的狀態，如紅血球數、白血球數、血小板數、血紅素含量、血球容積比等，醫師再根據這些檢驗數據做為臨床診斷的依據。目前，醫院多以侵入式的抽血檢查為主，送驗過程耗時且檢體在活體外可能受到環境因子影響導致變質進而影響檢查之正確性。 隨著科技的進步，非侵入式的醫學影像檢驗技術不斷推陳出新，在非線性光學顯微術中，倍頻式光學虛擬活體切片術不但有細胞等級的三維空間解析能力而且具有能量守恆的性質，不會對生物體造成傷害，檢測時亦無需使用染劑，因此，本研究團隊以人體手臂做為觀測點，透過手動方式調整物鏡及載物平移台形成三維控制，使雷射聚焦至皮下微血管進而觀測血球。由於微血管距離表皮僅200μm，檢測人員調整粗、細調節輪與載物平移台的力道過大或過小都容易使微血管被忽略，此外，微血管呈三維分布，很容易發生重複搜尋的情況導致效率變差，經驗不足的操作人員亦難上手，有鑒於此，本研究論文提出「自動搜尋皮下微血管演算法」輔助檢測人員找到微血管，如此不但能提升搜尋效率、減輕操作人員負擔更可避免與病患近距離接觸。Blood routine examination is important to check the health screening program, through the image of blood smear microscopy about the state of blood, such as red blood cells (RBCs), the white blood cells (WBCs), platelet count, hematocrit (Hct) are common examination in laboratory medicines, it requires a draw of blood, which is an invasive method to patients. It cost patients’ time to wait fo results and may be the subject of environmental factors on in vitro led todeterioration in turn affect the accuracy check. With the advancement of technology, non-invasive medical imaging inspection technology innovation, because of the nonlinear optical microscopy has sub-micron level tomography three-dimensional image resolution , has been widely used in biological observation and research. In the current nonlinear optical microscopy, harmonics-based in vivo optical virtual biopsy has three-dimensional cellular-level image, the system will not cause harm to the organism, but also without the use of stain .Therefore, the research team use this system for real-time observation and analysis of in vivo human blood cells, shape, flow and shape. We use arm as observation point, laser focus to subcutaneous capillaries. Before experiment, the participant loads his arm on the translation stage, operator adjusts the altitude of object and translation stage by manual, With three dimensional searching , program controls the laser focus to further observation of subcutaneous blood capillaries. As the capillaries from the epidermis only 200μm, operator adjustments coarse or fine adjustment knob of objective and controls translation stage powerfully , the capillary could be ignored, in addition, due to the capillary was three-dimensional distribution, the situation is prone to repeat the search resulted in poor efficiency, therefore, this thesis published “Algorithm of auto-searching subcutaneous capillaries”, not only can improve search efficiency, but also reduce the burden on the operator can be avoided to contact with patients.致謝 ..................................................I 中文摘要 .............................................III Abstract ............................................IV 目錄 ..................................................V 第一章 緒論 ............................................1 1.1背景 ...............................................1 1.2 斷層掃描技術 ........................................2 1.2.1光學同調斷層掃描 ................................5 1.2.2共軛焦顯微術 ...................................9 1.2.3倍頻式光學虛擬活體切片術 .........................12 1.3 研究動機 ...........................................14 第二章 基礎理論..........................................16 2.1光學倍頻顯微術 .......................................16 2.1.1二倍頻 ........................................16 2.1.2三倍頻 ........................................21 2.2流式細胞術 ..........................................22 2.3 皮膚系統 ...........................................25 2.4 數位影像處理 ........................................30 2.4.1 空間濾波器 ...................................30 2.4.2 平滑濾波器 ...................................32 2.4.3 排序濾波器 ...................................33 2.4.4 影像二值化 ...................................36 2.4.5直方圖 .......................................37 2.4.6直方圖拓寬 ....................................39 第三章 研究方法 .........................................41 3.1 光學系統 ...........................................41 3.2自動搜尋皮下微血管演算法 ...............................43 3.2.1流程圖 ........................................45 3.2.2搜尋步驟 ......................................46 第四章 實驗結果與討論 ....................................59 4.1倍頻影像剖面圖 .......................................60 4.2電腦搜尋結果 .........................................62 4.3人工搜尋結果 .........................................67 4.4搜尋時間之比較 .......................................74 第五章 結論與未來展望 ....................................76 參考文獻 ...............................................7714829287 bytesapplication/pdf論文公開時間：2016/11/20論文使用權限：同意無償授權倍頻顯微術自動搜尋演算法[SDGs]SDG3thesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/264443/1/ntu-102-R00548048-1.pdf