ThreeLayerSegmentation: Graph-Cut Pipeline for segmenting retinal layers

PRLEC Framework for OCT Processing and Visualization

class OCT_GUI.Algorithms.AutomaticSegmentation.ThreeLayerSegmentation.LayerSegmentation(volume, scaling=1, flattening_factor=4, statusText='')[source]

Class to segment three retinal layers of OCT images (ILM, RPE, BM).

The algorithm uses a graph-cut algorithm on each slice of the volume to segment the RPE boundary (similar to Chiu’s method). The result of the prior slice is used as input for the next slice, such that the resulting surface is smooth and accurate. The exponential weights are derived from the smoothed gradient images.

Methods

getResult() Running segmentation
run(mode) Running segmentation
runPipeline() Running pipeline for all layers
getResult()[source]

Running segmentation

Returns:
result: ndarray

final three layer segmentation

run(mode)[source]

Running segmentation

Parameters:
mode: string

‘ILM’ or ‘RPE’

runPipeline()[source]

Running pipeline for all layers

Returns:
result: ndarray

final three layer segmentation