|
|
Segmentation of retinal blood vessels in 3D images from optical coherence tomography
Žáková, Eliška ; Odstrčilík, Jan (referee) ; Štohanzlová, Petra (advisor)
The thesis deals with optical coherence tomography theory, eye and retinal structure and possibilities of retinal blood vessel segmentation in OCT scans. Further the segmentation of retinal layers used for blood vessels segmentation by OCTSEG application is shown. In the end 2D projections in the chosen layer of retinal blood vessels by different algorithms using mean, median, minimums and maximums in which are later segmented blood vessels are made and the results are compared. The best detection of blood vessels is in 2D projections made by algorithm using maximums. In the second case was better results of detection including background noise.
|
|
| |
|
|
Intra-retinal layer segmentation in optical coherence tomography images
Zlotý, Petr ; Kolář, Radim (referee) ; Gazárek, Jiří (advisor)
This work describes optical coherence tomography, it's principle and methods for automatic segmentation intra-retinal layers in OCT images . There is proposed algorithm for boundary detection of two intra retinal layers, RFNL and RPE layer based on thresholding and second algorithm for detection of seven layers, which is based on graph search algorithm. In summary is compared accuracy against manualy segmented layers and there is also proposed solution for problematic parts.
|
|
| |
|
|
Segmentation of retinal blood vessels in 3D images from optical coherence tomography
Žáková, Eliška ; Odstrčilík, Jan (referee) ; Štohanzlová, Petra (advisor)
The thesis deals with optical coherence tomography theory, eye and retinal structure and possibilities of retinal blood vessel segmentation in OCT scans. Further the segmentation of retinal layers used for blood vessels segmentation by OCTSEG application is shown. In the end 2D projections in the chosen layer of retinal blood vessels by different algorithms using mean, median, minimums and maximums in which are later segmented blood vessels are made and the results are compared. The best detection of blood vessels is in 2D projections made by algorithm using maximums. In the second case was better results of detection including background noise.
|
|
|
Intra-retinal layer segmentation in optical coherence tomography images
Zlotý, Petr ; Kolář, Radim (referee) ; Gazárek, Jiří (advisor)
This work describes optical coherence tomography, it's principle and methods for automatic segmentation intra-retinal layers in OCT images . There is proposed algorithm for boundary detection of two intra retinal layers, RFNL and RPE layer based on thresholding and second algorithm for detection of seven layers, which is based on graph search algorithm. In summary is compared accuracy against manualy segmented layers and there is also proposed solution for problematic parts.
|
guest ::
