Automated thresholding method (ATM) for biomass fraction determination using FISH and confocal microscopy

BACKGROUND: Fluorescence in situ hybridization (FISH) together with confocal laser scanning microscopy (CLSM) is widely used for the analysis and quantification of biomass fractions of activated sludge or biofilm systems in wastewater treatment research. Unlike the FISH technique, the CLSM image analysis and quantification is not generally unique and thus, several manual and automated methods exist that can lead to very different results. RESULTS: A method based on the analysis of negative control images is reported in this paper. Histograms of negative control images were used in the segmentation of positive images (hybridized with FISH probes). The threshold (I-thr) was defined as the minimum intensity value that satisfies the following condition: the proportion of pixels with intensity I <= I-thr in the negative control images is greater than or equal to a value alpha. It was found that the optimal alpha-value was 99.90% for the specific case studied (nitrifying activated sludge samples) by comparing the automatic thresholding with manual thresholding by three expert operators. CONCLUSION: The method developed was useful for following systematically the evolution of ammonia oxidizing bacteria (AOB) and nitrite oxidizing bacteria (NOB) fractions in a nitrifying system under changing operating conditions. (C) 2009 Society of Chemical Industry