GraphWeights: ILM/RPE Graph-Weight calculation for Three layer Segmentation

PRLEC Framework for OCT Processing and Visualization

OCT_GUI.Algorithms.AutomaticSegmentation.GraphWeights.calculateGradients(_slice, mode, BF_bscan)[source]

Helper to calculate gradients

Parameters:
_slice: numpy array 2D

oct input volume slice

BF_bscan: ndarray, int, size 3

Bilteral filter values for smoothing

OCT_GUI.Algorithms.AutomaticSegmentation.GraphWeights.runWeightCalculation(volume, mode, dictParameters)[source]

Calculating weights for modes

The volume is bilateral filter in the en-face plane. Next, each B-scan is bilateral filtered again.

RPE: The result of this is returned as weights ILM: The gradient is calculated (Dark-to-bright) and the resulting slice exposured to stretch the grayvalues to 0…1.

Parameters from parameters:
  • AUTO_ILM_BF_ENFACE : En face plane smoothing values for ILM
  • AUTO_ILM_BF_BSCAN : B-scan smoothing values for ILM
  • AUTO_RPE_BF_ENFACE : En face plane smoothing values for RPE
  • AUTO_RPE_BF_BSCAN : B-scan smoothing values for RPE
Parameters:
volume: numpy array 3D

oct input volume

mode: string

‘RPE’ or ‘ILM’ mode setter to detect filter values

dictParameters: dictionary

Parameters from parameters.txt