Sinead Healy1, Jill McMahon2, Peter Owens3, Peter Dockery4, Una FitzGerald5. 1. Galway Neuroscience Centre, School of Natural Sciences, Biomedical Sciences Building, National University of Ireland, Galway, Ireland. Electronic address: Sinead.healy@nuigalway.ie. 2. Galway Neuroscience Centre, School of Natural Sciences, Biomedical Sciences Building, National University of Ireland, Galway, Ireland. Electronic address: Jill.mcmahon@nuigalway.ie. 3. Centre for Microscopy and Imaging, National University of Ireland, Galway, Ireland. Electronic address: Peter.owens@nuigalway.ie. 4. Centre for Microscopy and Imaging, National University of Ireland, Galway, Ireland. Electronic address: Peter.dockery@nuigalway.ie. 5. Galway Neuroscience Centre, School of Natural Sciences, Biomedical Sciences Building, National University of Ireland, Galway, Ireland. Electronic address: Una.fitzgerald@nuigalway.ie.
Abstract
BACKGROUND: Image segmentation is often imperfect, particularly in complex image sets such z-stack micrographs of slice cultures and there is a need for sufficient details of parameters used in quantitative image analysis to allow independent repeatability and appraisal. NEW METHOD: For the first time, we have critically evaluated, quantified and validated the performance of different segmentation methodologies using z-stack images of ex vivo glial cells. The BioVoxxel toolbox plugin, available in FIJI, was used to measure the relative quality, accuracy, specificity and sensitivity of 16 global and 9 local threshold automatic thresholding algorithms. RESULTS: Automatic thresholding yields improved binary representation of glial cells compared with the conventional user-chosen single threshold approach for confocal z-stacks acquired from ex vivo slice cultures. The performance of threshold algorithms varies considerably in quality, specificity, accuracy and sensitivity with entropy-based thresholds scoring highest for fluorescent staining. COMPARISON WITH EXISTING METHODS: We have used the BioVoxxel toolbox to correctly and consistently select the best automated threshold algorithm to segment z-projected images of ex vivo glial cells for downstream digital image analysis and to define segmentation quality. The automated OLIG2 cell count was validated using stereology. CONCLUSIONS: As image segmentation and feature extraction can quite critically affect the performance of successive steps in the image analysis workflow, it is becoming increasingly necessary to consider the quality of digital segmenting methodologies. Here, we have applied, validated and extended an existing performance-check methodology in the BioVoxxel toolbox to z-projected images of ex vivo glia cells.
BACKGROUND: Image segmentation is often imperfect, particularly in complex image sets such z-stack micrographs of slice cultures and there is a need for sufficient details of parameters used in quantitative image analysis to allow independent repeatability and appraisal. NEW METHOD: For the first time, we have critically evaluated, quantified and validated the performance of different segmentation methodologies using z-stack images of ex vivo glial cells. The BioVoxxel toolbox plugin, available in FIJI, was used to measure the relative quality, accuracy, specificity and sensitivity of 16 global and 9 local threshold automatic thresholding algorithms. RESULTS: Automatic thresholding yields improved binary representation of glial cells compared with the conventional user-chosen single threshold approach for confocal z-stacks acquired from ex vivo slice cultures. The performance of threshold algorithms varies considerably in quality, specificity, accuracy and sensitivity with entropy-based thresholds scoring highest for fluorescent staining. COMPARISON WITH EXISTING METHODS: We have used the BioVoxxel toolbox to correctly and consistently select the best automated threshold algorithm to segment z-projected images of ex vivo glial cells for downstream digital image analysis and to define segmentation quality. The automated OLIG2 cell count was validated using stereology. CONCLUSIONS: As image segmentation and feature extraction can quite critically affect the performance of successive steps in the image analysis workflow, it is becoming increasingly necessary to consider the quality of digital segmenting methodologies. Here, we have applied, validated and extended an existing performance-check methodology in the BioVoxxel toolbox to z-projected images of ex vivo glia cells.
Authors: Karolin Engelkes; Jennice Helfsgott; Jörg U Hammel; Sebastian Büsse; Thomas Kleinteich; André Beerlink; Stanislav N Gorb; Alexander Haas Journal: J Anat Date: 2019-05-07 Impact factor: 2.610
Authors: Aris I Giotakis; Jozsef Dudas; Rudolf Glueckert; Daniel Dejaco; Julia Ingruber; Felix Fleischer; Veronika Innerhofer; Leyla Pinggera; Ljilja Bektic-Tadic; Sarah A M Gabriel; Herbert Riechelmann Journal: Histochem Cell Biol Date: 2020-11-29 Impact factor: 4.304
Authors: Amy Keerie; Heledd Brown-Wright; Isaac Kirkland; Andrew Grierson; James J P Alix; Christian Holscher; Richard J Mead Journal: Sci Rep Date: 2021-08-23 Impact factor: 4.379
Authors: Lauren M Stein; Rinzin Lhamo; Anh Cao; Jayme Workinger; Ian Tinsley; Robert P Doyle; Harvey J Grill; Gerlinda E Hermann; Richard C Rogers; Matthew R Hayes Journal: Transl Psychiatry Date: 2020-03-09 Impact factor: 6.222