Application of Active Contour Models on Medical Diagnosis
| Autor: | Ying-Yao Tu, 塗英堯 |
|---|---|
| Rok vydání: | 2009 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 97 Glaucoma is one of blind diseases in the ophthalmology, and treatments decelerate worse situation only can temporarily. There is no way to prevent glaucoma, therefore early discovery and treatment is important. The Auto Fundus Camera (Fundus) and Optical Coherence Tomography (OCT) are two medical equipment to assess glaucoma disease. The study discusses disc and cup characteristic in color Fundus images. This study is to apply Active Contour Models (ACM) to capture the features of optic disc in Fundus, and to detect the location and size of the optic disc. In addition, the results from ACM will be further studied with the parameters by ophthalmologist’s manual segmentation and OCT results. The fundus image will be. First, carried out for the original image contrast adjustment. Second use the RGB separation technology and its image is divided into R, G and B images. Third, one of the R, G and B images will be selected for further analysis. ACM is used capture the disc and cup, and then the cup-to-disc area ratio (C/D area ratio) is calculated. This study uses 8 Normal cases and 12 glaucoma cases. Correlation analysis is applied on the results from ACM, ophthalmologist and OCT to test results for Correlation analysis. The correlation coefficient between normal database and OCT is 0.38; and the correlation coefficient between normal database and ophthalmologist is 0.81. The correlation coefficient between glaucoma and OCT is 0.55. The results show that there is moderate correlation between normal and glaucomatous cases by using ACM. Besides, there is high consistency among our method, OCT and ophthalmologists. The research provides good information to assist ophthalmologists on clinical examination. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|