Hierarchical Mergence Approach to Cell Detection in Phase Contrast Microscopy Images

Abstract

Phase contrast microscope is one of the most universally used instruments to observe long-term cell movements in different solutions. Most of classic segmentation methods consider a homogeneous patch as an object, while the recorded cell images have rich details and a lot of small inhomogeneous patches, as well as some artifacts, which can impede the applications. To tackle these challenges, this paper presents a hierarchical mergence approach (HMA) to extract homogeneous patches out and heuristically add them up. Initially, the maximum region of interest (ROI), in which only cell events exist, is drawn by using gradient information as a mask. Then, different levels of blurring based on kernel or grayscale morphological operations are applied to the whole image to produce reference images. Next, each of unconnected regions in the mask is applied with Otsu method independently according to different reference images. Consequently, the segmentation result is generated by the combination of usable patches in all informative layers. The proposed approach is more than simply a fusion of the basic segmentation methods, but a well-organized strategy that integrates these basic methods. Experiments demonstrate that the proposed method outperforms previous methods within our datasets.

Deep Learning Automates the Quantitative Analysis of Individual Cells in Live-Cell Imaging Experiments

“…As a result, the cytoplasmic segmentation problem is frequently circumvented by sampling pixels in close proximity to the nucleus and using them as a proxy for the cytoplasm [27–29]. More progress has been made in segmenting closely packed bacterial cells [17]; however, a robust method to identify the cytoplasm of mammalian cells or bacterial micro-colonies with single-cell resolution directly from phase microscopy images has remained elusive [17, 26, 30, 31]. …”

Detection of Dendritic Spines Using Wavelet-Based Conditional Symmetric Analysis and Regularized Morphological Shared-Weight Neural Networks

“…Due to the complexity of the dendrites and dendrite spines' distribution, we have employed morphological operation to remove the small isolated point-set for the dendrite in the binary image obtained by local Otsu [ 27 – 29 ] via ( 1 ), which could decrease the disconnection rate of the dendrite detection: in which n is the threshold of the number of positive pixels. The value of n could be determined by trial and error method and means that the pixel belongs to the major line if there are more than n positive pixels in its 3 × 3 neighborhood window.…”

Machine-Learning-Aided Quantification of Area Coverage of Adherent Cells from Phase-Contrast Images

“…Phase contrast is superior in that it allows imaging of the cells without any invasive staining operation, however it is not quantitative in the sense explained above. Owing to this fact, much work has been invested in quantifying phase-contrast images over the years (Ambühl et al, 2012; Theriault et al, 2012; Yin et al, 2012; Su et al, 2013; Chen et al, 2014; Jaccard et al, 2014).…”