Artificial neural network-aided image analysis system for cell counting

Abstract

Background: In histological preparations containing debris and synthetic materials, it is difficult to automate cell counting using standard image analysis tools, i.e., systems that rely on boundary contours, histogram thresholding, etc. In an attempt to mimic manual cell recognition, an automated cell counter was constructed using a combination of artificial intelligence and standard image analysis methods. Methods: Artificial neural network (ANN) methods were applied on digitized microscopy fields without pre‐ANN feature extraction. A three‐layer feed‐forward network with extensive weight sharing in the first hidden layer was employed and trained on 1,830 examples using the error back‐propagation algorithm on a Power Macintosh 7300/180 desktop computer. The optimal number of hidden neurons was determined and the trained system was validated by comparison with blinded human counts. System performance at 50x and 100x magnification was evaluated. Results: The correlation index at 100x magnification neared person‐to‐person variability, while 50x magnification was not useful. The system was approximately six times faster than an experienced human. Conclusions: ANN‐based automated cell counting in noisy histological preparations is feasible. Consistent histology and computer power are crucial for system performance. The system provides several benefits, such as speed of analysis and consistency, and frees up personnel for other tasks. Cytometry 36:18–26, 1999. © 1999 Wiley‐Liss, Inc.

Unbiased estimation of cell number using the automatic optical fractionator

“…A survey of segmentation algorithms proposed to improve efficiency of cell counting methods show these segmentation approaches focus on finding number of 2-D cell profiles [Sjöström et al, 1999; Nattkemper et al, 2001; Ray et al, 2002; Benali et al, 2003; Peng et al, 2003; Lin et al, 2005; Long et al, 2005, 2006; Costa & Bollt 2006; Inglis et al, 2008; Ho et al, 2011; Liu et al, 2014; de Gracia, et al, 2015]. Per the Delesse principle (1847), the total number of arbitrary 3-D cells on tissue sections is not equal to the total number of their 2-D profiles, i.e., Total N cells ≠ Total N profiles .…”

Cryo-imaging of fluorescently labeled single cells in a mouse

“…Typical cellular quantification algorithms range in sophistication from simple manual dot counting, to deformable model [11] and neural network based segmentation [12]. All of these methods require images of sufficient resolution, such that the cell boundary is distinguishable.…”

A Deep Learning-Based System (Microscan) for the Identification of Pollen Development Stages and Its Application to Obtaining Doubled Haploid Lines in Eggplant

“…The main issue with current automatized systems for the identification of the stages of development of pollen is the approach used. Generally, this is based solely on trying to parameterize, from the expert knowledge, the complexity of the image of a cell [20][21][22]. This approach has largely failed to date, because the number of variables that must be parameterized for this approach to work successfully is very large.…”