Journal of Biomedical Optics · 2018 · 11 citations · 19 references
Natural SciencesBiomedical EngineeringCell BiologyBiomedical ImagingBiomedical OpticsMedical VisualizationCell StructureCellular ImagingConfocal MicroscopyFluorescence MicroscopyLight MicroscopyMicroscopic Cell Morphology3D Cell ReconstructionCell Morphology AnalysisCell Morphology AssessmentConfocal Fluorescence MicroscopyOptical NoiseScatteringThree-dimensional Cell ImagingThree-dimensional Optical Imaging
Constructing models of cells' realistic internal and external morphology is vital for correlation between light scattering and morphology of the scattering structure. The image stack obtained from fluorescent confocal microscopy is at present used to construct the cell's three-dimensional (3-D) morphology. However, due to the poor labeling quality and unavoidable optical noise present in the image stacks, 3-D morphologies are difficult to construct and are an impediment to the statistical analyses of cell structures. We propose a method called the "area and shape constraint method (ASCM)" for constructing 3-D morphology. Blurred 3-D morphologies constructed by common methods from image stacks considered as defective and which are commonly discarded are well restored by the ASCM. Seventy-four clinical blood samples and a series of standard fluorescent spheres are selected to evaluate the validity and precision of our proposed ASCM. Both the qualitative and quantitative results obtained by ASCM indicate the good performance of the method in constructing the cell's 3-D morphology.
19
Deep Learning in Label-free Cell Classification
Claire Lifan Chen, Ata Mahjoubfar, Li‐Chia Tai et al. · Scientific Reports · 2016 · 472 citations · Full text
R. Scott Brock, Xin‐Hua Hu, Douglas A. Weidner et al. · Journal of Quantitative Spectroscopy and Radiative Transfer · 2006 · 62 citations