Automated quantification and sizing of unbranched filamentous cyanobacteria by model based object oriented image analysis

Abstract

The quantification and sizing of filamentous cyanobacteria in environmental samples or cultures is a time consuming process often performed by manual or semi-automated microscopic analysis. Automation by conventional image analysis is difficult because filaments may exhibit large variation in length and patchy autofluorescence. Moreover, individual filaments frequently cross each other on microscopic preparations, as can be deduced by modelling. This study presents a novel approach based on object oriented image analysis to simultaneously retrieve (a) filament number, (b) individual filament lengths, and (c) cumulative filament length of unbranched cyanobacterial morphotypes on fluorescence microscopic images in a fully automated high-throughput manner. Special emphasis was put on the correct detection of overlapping objects by image analysis and on the appropriate coverage of filament length distribution by using large composite images. The method was validated on a dataset of Planktothrix rubescens from field samples and compared with manual filament tracing, the line intercept method, and the Utermöhl counting approach. The described computer program features batch processing of large images from any appropriate source and annotation of detected filaments. It requires no user interaction, is freely available and might thus be a useful tool for basic research and drinking water quality control.