Abstract:Single cell segmentation is critical and challenging in live cell imaging data analysis. Traditional image processing methods and tools require time-consuming and labor-intensive efforts of manually fine-tuning parameters. Slight variations of image setting may lead to poor segmentation results. Recent development of deep convolutional neural networks(CNN) provides a potentially efficient, general and robust method for segmentation. Most existing CNN-based methods treat segmentation as a pixel-wise classification problem. However, three unique problems of cell images adversely affect segmentation accuracy: lack of established training dataset, few pixels on cell boundaries, and ubiquitous blurry features. The problem becomes especially severe with densely packed cells, where a pixel-wise classification method tends to identify two neighboring cells with blurry shared boundary as one cell, leading to poor cell count accuracy and affecting subsequent analysis. Here we developed a different learning strategy that combines strengths of CNN and watershed algorithm. The method first trains a CNN to learn Euclidean distance transform of binary masks corresponding to the input images. Then another CNN is trained to detect individual cells in the Euclidean distance transform. In the third step, the watershed algorithm takes the outputs from the previous steps as inputs and performs the segmentation. We tested the combined method and various forms of the pixel-wise classification algorithm on segmenting fluorescence and transmitted light images. The new method achieves similar pixel accuracy but significant higher cell count accuracy than pixel-wise classification methods do, and the advantage is most obvious when applying on noisy images of densely packed cells.

Cell image separation based on shortest distance

Silhouette coefficient based approach on cell-phone classification for unknown source images

An Algorithm to Separate Overlapping Particles Based on Connected Operator

A New Algorithm for Separating Cell in Bacteria Image

Finding splitting lines for touching cell nuclei with a shortest path algorithm

Contour separation based on the feature clustering of edges

Clustered nuclei splitting via curvature information and gray-scale distance transform

Interactive Color Separation and Replacement for Color Image

A novel geodesic distance based clustering approach to delineating boundaries of touching cells

A Separating Algorithm for Overlapping Red Blood Cell Images

Automatic microarray image segmentation with clustering-based algorithms

A New Approach to Separating Touching Spots in Particle Images

A Method of Automatic Cell Counting Based on Microscopic Image

Method of Automatic Brain Tumor Segmentation Based on Clustering Tree Matching

Morphologically Restricted B-snake Model for Clustering It Cell Image Segmentation

Cell Segmentation Via Spectral Analysis on Phase Retardation Features

Image Segmentation with Mean Shift and Region Merging Methods

Ensemble-based markov random field clustering for colony image analysis

Clumped Nuclei Segmentation with Adjacent Point Match and Local Shape based Intensity Analysis for Overlapped Nuclei in Fluorescence In-Situ Hybridization Images

Learn to segment single cells with deep distance estimator and deep cell detector

A Cell Marker-Based Clustering Strategy (cmcluster) for Precise Cell Type Identification of Scrna-Seq Data