Determination of cell number and size of a population of Pseudomonas fluorescens by image analysis

Abstract

Image analysis is a very useful technique for counting and sizing bacteria, minimising hum an operati on and providing accurate results in a short inte rval of time. Microscopic observations of a population of Pseudomonas fluo rescens were digitised by a frame grabber and the grabbed images were enhanced by background subtraction and multiplication of two copies. To extract the objects from the background, an appropriate threshold had to be chosen. Full grown single bacterial cells showed to be normally distributed around two mean sizes, one corresponding to standi ng bacteria and the other to lying bac teria. Two Gauss functions were least square fitted to the se data points resulting in the mean area and the standard deviation. The enumeration of single cells was obtai ned from the area of each gauss curve. It was also possible to determ ine the number of single bacteria in aggregates, once the mean project area of a single cell is known. The enumeration was made for each threshold selected. The number of particles coun ted was constant in a large range of threshold. whereas the cell area increases with the threshold ins talled.