The microscope system automatically identifies and counts bacteria (shown here circled).
Microscope solution for counting bacteria
Preventative Health Flagship researchers have developed an integrated microscope system to automatically count bacteria in human faeces.
1 November 2007 | Updated 15 October 2012
Meeting the need
Scientists with the Preventative Health Flagship are searching for bacterial species in human faeces that may protect against colorectal cancer.
The research requires fast and accurate counts of fluorescently tagged bacteria from faecal samples.
Two image scientists and a biochemist from the Flagship set about creating an automated fluorescence microscope system that would achieve greater speed and accuracy than manual counting.
The scientists started with a fluorescence microscope and a slide loader and wrote software to automate the image acquisition.
To resolve problems with contaminating matter and clumping, they developed a proprietary method for preparing the faecal samples that ensures bacteria of interest are evenly distributed on the microscope slide and are clearly visible under the microscope.
Comparison with human operators shows the automated microscope system does its job in a fraction of the time.
They wrote algorithms that:
automatically focus the microscope
determine which parts of the image represent bacteria
count the bacteria and record the results in a spreadsheet.
To enable completely unattended operation, the system contains a custom designed auto-focus sub-system that:
Comparison with human operators shows the automated microscope system does its job well and in a fraction of the time.
It could potentially be used in many other areas requiring analysis of high volumes of microscopy samples.
Commercial applications are being explored.
This work involved CSIRO imaging specialists Dr Paul Jackway and Dr Volker Hilsenstein and biochemist Dr Michael Conlon.
Find out more about the work of the Preventative Health Flagship.
Hilsenstein V. 2005. Robust Autofocusing for Automated Microscopy Imaging of Fluorescently Labelled Bacteria. In: Proceedings of the Digital Image Computing: Techniques and Applications. Cairns, Queensland, Australia. 6-8 December 2005. IEEE Computer Society Press. (DICTA 2005). Pp. 95-101.