Autofluorescence spectrum and excitation-emission matrices for cervical pre-cancer tissue differentiation.
| Autor: | Jui-Wen Teng, 鄧睿玟 |
|---|---|
| Rok vydání: | 2003 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 91 In recent year, thousands women died cervical cancer in Taiwan. In all of the universal diagnosis, Pap smear is the more convenient but has higher false negative. Although Colposcopy has higher accuracy, it can’t detect lesions at the cervix that shrink inside. Light-induced autofluorescence detects the fluorophores, such as collagen, NADH, FAD, in the biological tissues. Fluorophores represent different spectrum along with the development of normal tissues to abnormal tissues to early diagnose the disease. Light-induced autofluorescence is low invasive, early operating, real time, and highly potential optical biopsy technique for precancer diagnosis. We utilized this light-induced autofluorescence technique for cervical cancer study and measured normal and abnormal tissues (cervical intraepithelial neoplasia, squamous cell carcinoma). Used PLS (partial least square) to discriminate between normal and CIN tissue. Our study present the PLS is an optimum algorithm in tissues disease diagnosis. The EEMs provides a tool with which to study the relative diagnostic histology of tissue. We utilized two dimensions partial least squares (2D-PLS) to analysis EEMs data. These results showed that 2D-PLS had both high sensitivity, specificity and accuracy. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|