An Automated Method for Cell Detection in Zebrafish

Abstract

Quantification of cells is a critical step towards the assessment of cell fate in neurological disease or developmental models. Here, we present a novel cell detection method for the automatic quantification of zebrafish neuronal cells, including primary motor neurons, Rohon-Beard neurons, and retinal cells. Our method consists of four steps. First, a diffused gradient vector field is produced. Subsequently, the orientations and magnitude information of diffused gradients are accumulated, and a response image is computed. In the third step, we perform non-maximum suppression on the response image and identify the detection candidates. In the fourth and final step the detected objects are grouped into clusters based on their color information. Using five different datasets depicting zebrafish cells, we show that our method consistently displays high sensitivity and specificity of over 95%. Our results demonstrate the general applicability of this method to different data samples, including nuclear staining, immunohistochemistry, and cell death detection.

Automatic Identification of Fluorescently Labeled Brain Cells for Rapid Functional Imaging

“…Both steps use classifiers generated by the RobustBoost algorithm (Freund 2009), which is relatively insensitive both to explicit errors in human annotations and inconsistencies in labeling of ambiguous regions (Schapire et al 1998). RobustBoost is part of a family of machine learning algorithms called Boosting, which have been used in several biological image segmentation problems (Giannone et al 2007;Liu et al 2008). The classifiers consist of a nonbinary decision tree whose nodes correspond to thresholds on selected features and whose output is a score that corresponds to whether a given pixel is part of a cell.…”

Automated three-dimensional detection and counting of neuron somata

“…In contrast, threshold-or watershed-based approaches display relatively better detection sensitivity, i.e. they find all objects, but usually result in incorrect numbers (Liu et al, 2008), e.g. for our sample data touching, densely clustered neurons are counted as one, resulting in about 20% less neurons.…”

“…The obvious disadvantage of manual neuron detection, apart from possible subjectivity, is the amount of time needed for neuron counting. In consequence, automated accurate detection and segmentation of neurons from microscopic images has been extensively studied (Liu et al, 2008). In general, these algorithms can be divided into three categories: threshold-based (Wu et al, 2000;Wu et al, 1995), watershedbased (Lin et al, 2003;Lin et al, 2005;Malpica et al, 1997;Nilsson and Heyden, 2005;Vincent and Soille, 1991) and model-based approaches (Chang and Parvin, 2006;Li et al, 2006;Lin et al, 2007;Lin et al, 2005;Raman et al, 2007;Ranzato et al, 2007).…”

An Automated Three-Dimensional Detection and Segmentation Method for Touching Cells by Integrating Concave Points Clustering and Random Walker Algorithm

“…Gradient flow tracking, another extension of the deformed model method, has been proposed to segment touching cells. However, it is sensitive to heterogeneous brightness [3], [14]–[16], which may lead to inaccurate flow values and error direction, especially inside the cell where the gradient may not flow toward the cell center. More recently, researchers have introduced multi-scale LoG filtering which achieved good results for DAPI-stained slices [3].…”