Livelayer

Livelayer performs semi-automatic segmentation of Optical Coherence Tomography (OCT) images to delineate eight macular layers and detect retinal pathologies such as diabetic macular edema.


Key Features:

  • Graph-based segmentation: Semi-automatic graph-based segmentation integrating Dijkstra’s Shortest Path (SPF) algorithm with the Live-wire function for retinal layer and fluid boundary detection.
  • Layer coverage: Segmentation of eight distinct macular layers, with explicit references to layers such as ILM, IPL-INL, and OPL-ONL.
  • Fluid detection: Detection and segmentation of retinal fluids, including both clear and unclear fluid objects.
  • Preprocessing operations: Image preprocessing steps to prepare OCT scans for segmentation.
  • Time efficiency: Reported per-layer processing times from approximately 5 seconds for ILM to 15.57 seconds for OPL-ONL.
  • Accuracy and concordance: Unsigned segmentation errors averaging 1.9–2.7 pixels for layers (ILM, IPL-INL, OPL-ONL) with Bland–Altman analysis indicating concordance with manual segmentation.
  • Repeatability: Repeatability error around one pixel for the OPL-ONL layer and less than one pixel for other layers.
  • Irregularity index: Calculation of an irregularity index in deformed OCT images for structural analysis.
  • Pathology identification: Identification of retinal pathologies such as diabetic macular edema from OCT-derived layer and fluid segmentation.

Scientific Applications:

  • Ophthalmic disease research and diagnosis: Analysis of macular layer morphology and detection of diabetic macular edema in OCT images for research and clinical assessment.
  • Retinal fluid quantification: Identification and segmentation of retinal fluids to support quantitative studies of edema and fluid-related pathology.
  • Structural deformation analysis: Use of the irregularity index to characterize deformed OCT images in studies of retinal deformities and related neurological disorders detectable by OCT.

Methodology:

Graph-based semi-automatic segmentation integrating Dijkstra’s Shortest Path (SPF) algorithm with the Live-wire function, OCT image preprocessing operations, segmentation of eight macular layers and retinal fluids, and irregularity index computation.

Topics

Details

Tool Type:
command-line tool, desktop application
Programming Languages:
MATLAB
Added:
3/19/2021
Last Updated:
4/11/2021

Operations

Publications

Montazerin M, Sajjadifar Z, Pour EK, Riazi-Esfahani H, Mahmoudi T, Rabbani H, Movahedian H, Dehghani A, Akhlaghi M, Kafieh R. Livelayer: A Semi-Automatic Software Program for Segmentation of Layers and Diabetic Macular Edema in Optical Coherence Tomography Images. Unknown Journal. 2021. doi:10.21203/rs.3.rs-246833/v1.