348
L. Pan and X. Chen
Fig. 12.2 Flowchart of the proposed system
12.3.1 Initialization
An initial segmentation of the SEAD regions is required to initialize the graph-based
SEAD segmentation algorithm. A statistical voxel classification approach is applied
directly to the preprocessed input image to find voxels that are likely inside of a
SEAD region. The classifier assigns a likelihood to each voxel that it belongs to a
SEAD. This likelihood map serves as constraints for the graph-based segmentation
algorithm.
12.3.1.1 Preprocessing
First, the output of our 11-surface segmentation [29] is used to determine the upper
and lower surface of the retina in the scan. For these 11 surfaces, the top retinal surface
corresponds to the inner limiting membrane and bottom layer corresponds to the RPE.
Although, the top retinal surface is usually segmented successfully even in OCT scans
L. Pan and X. Chen
Fig. 12.2 Flowchart of the proposed system
12.3.1 Initialization
An initial segmentation of the SEAD regions is required to initialize the graph-based
SEAD segmentation algorithm. A statistical voxel classification approach is applied
directly to the preprocessed input image to find voxels that are likely inside of a
SEAD region. The classifier assigns a likelihood to each voxel that it belongs to a
SEAD. This likelihood map serves as constraints for the graph-based segmentation
algorithm.
12.3.1.1 Preprocessing
First, the output of our 11-surface segmentation [29] is used to determine the upper
and lower surface of the retina in the scan. For these 11 surfaces, the top retinal surface
corresponds to the inner limiting membrane and bottom layer corresponds to the RPE.
Although, the top retinal surface is usually segmented successfully even in OCT scans