Assessing the efficacy of low-level image content descriptors for computer-based fluorescence microscopy image analysis
Author:
- 23 May 2011
- journal article
- Published by Wiley in Journal of Microscopy
- Vol. 243 (3) , 284-292
Abstract
The increasing prevalence of automated image acquisition systems and state‐of‐the‐art information technology has enabled new types of microscopy experiments based on automatic processing of massive image data sets, and numerous methods of high‐content screening using machine vision and pattern recognition methods have been proposed. However, as a relatively young discipline, it is important to validate these methods and ensure that the machine vision and pattern recognition techniques reliably reflect the actual morphology, and can be effectively used for finding and validating scientific discoveries. In this report we show that some of the previously reported experimental results using automatic microscopy image analysis might be biased, and discuss practices and methods that can be used to obtain objective and reliable automatic analysis of microscopy images.Keywords
This publication has 35 references indexed in Scilit:
- Computational prediction of neural progenitor cell fatesNature Methods, 2010
- Principles of Bioimage Informatics: Focus on Machine Learning of Cell Patterns2010
- Extracting Rich Information from Images2009
- Image-based chemical screeningNature Chemical Biology, 2007
- One-shot learning of object categoriesIEEE Transactions on Pattern Analysis and Machine Intelligence, 2006
- Boosting accuracy of automated classification of fluorescence microscope images for location proteomicsBMC Bioinformatics, 2004
- Systematic genome-wide screens of gene functionNature Reviews Genetics, 2004
- Integrated Optomechanical Analysis2002
- A neural network classifier capable of recognizing the patterns of all major subcellular structures in fluorescence microscope images of HeLa cellsBioinformatics, 2001
- Automated recognition of patterns characteristic of subcellular structures in fluorescence microscopy imagesCytometry, 1998