Skip to main content

Table 1 List of tested segmentation methods and all-in-one segmentation tools and definition of abbreviations

From: Cell segmentation methods for label-free contrast microscopy: review and comprehensive comparison

Segmentation step Abbreviation Description Setable parameters Computational time Ref.
All in one tools
  aioFasright Nucleus editor of Farsight toolkit N/A 4.96 s [2]
  aioCellX segmentation, fluorescence quantification, and tracking tool CellX N/A 10.30 s [3]
  aioFogbank single cell segmentation tool FogBank according Chalfoun N/A 12.00 s [4]
  aioFastER fastER - user-friendly tool for ultrafast and robust cell segmentation N/A 0.42 s [5]
  aioCellProfiler tool for cell analysis pipelines including single cell segmentation N/A 11.8 s [10]
  aioDMGW Dry mass-guided watershed method, (Q-PHASE, Tescan)   1.00 s  
Reconstruction
  rDIC-Koos DIC/HMC image reconstruction according Koos 2 36.60 min [12]
  rDIC-Yin DIC/HMC image reconstruction according Yin 2 2.10 s [13]
  rPC-Yin PC image reconstruction according Yin 4 13.10 min [14]
  rPC-Tophat PC image reconstruction according Thirusittampalam and Dewan 1 0.17 s [15, 16]
Foreground-background segmentation
  sST simple thresholding 1 <0.01 s  
  sOtsu thresholding using Gaussian distribution 0 <0.01 s [17]
  sPT thresholding using Poisson distribution 0 <0.01 s [2]
  sEGT empirical gradient threshold 3 0.24 s [18]
  sPC-Juneau Feature extraction approach according Juneau 3 0.26 s [19]
  sPC-Topman Feature extraction approach according Topman 4 0.35 s [20]
  sPC-Phantast Phantast toolbox developed by Jaccard 5 0.35 s [21]
  sLS-Caselles Level-set with edge-based method 2 31.40 s [22]
  sLS-ChanVese Level-set with region-based method 2 11.10 s [23]
  sGraphCut Graph-Cut applied on recosntructed and raw data 2 15.80 s [24]
  sWekaGraphCut Graph-Cut applied on probability maps generated by Weka 2 31.80 min** [24]
  sIlastikGraphCut Graph-Cut applied on probability maps generated by Ilastik 2 31.52 min** [24]
  sIlastik machine learning tool by Sommer N/A 31.20 min+21 s* [25].
  sWeka machine learning tool by Arganda-Carreras N/A 27.60 min+2.20 min* [26]
Cell detection (seed-point extraction)
  dLoGm-Peng multiscale LoG by Peng 4 3.60 s [27]
  dLoGm-Kong multiscale LoG by Kong 5 4.20 s [28]
  dLoGg-Kong generalized LoG filter by Kong 2 46.40 s [28]
  dLoGg-Xu generalized LoG filter by Xu 3 5.10 s [29]
  dLoGh-Zhang Hessian analysis of LoG images by Zhang 1 8.90 s [30]
  dFRST fast radial-symmetry transform 5 153.10 s [31]
  dGRST generalized radial-symmetry transform 5 572.30 s [32]
  dRV-Qi radial voting methods by Qi et al. 5 95.00 s [33]
  dDT-Threshold distance transform by Thirusittampalam, threshold-generated foreground 4 0.11 s [15]
  dDT-Weka distance transform by Thirusittampalam, sWeka-generated foreground 3 0.11 s [15]
  dMSER maximally stable extremal region method (MSER) 3 2.10 s [34]
  dCellDetect machine learning method based on MSER 1 141.70 s/60.20 s* [35]
Single cell (instance) segmentation
  MCWS Marker-conttrolled watershed 0 1.40 s  
  MCWS-dDT Marker-conttrolled watershed on DT image 0 1.41 s  
  1. For detailed list of optimized parameters see Additional file 1. * computational time for learning based approaches indicated as two values for learning and classification. ** computational time for Weka+Graph cut combination shown as sum time of these methods. not includes time for Weka probability map creation, indicate final segmentation step following foreground-background segmentation and seed-point extraction. Number of parameters in “all-in-one” approaches not shown because of the GUI-based nature, similarly, not shown for learning-based approaches, see Methods section for details. Computational time shown for one 1360 ×1024 DIC field of view