Digital color correlation for the recognition of Vibrio cholerae 01 in laboratory and environmental samples

In general, direct counts are more reflective of microbial abundance than plate counts. Microbiological techniques developed to recognize and count microorganisms in natural systems have had problems with reliability. Different methods frequently give different results for the enumeration of a specific organism. For direct microscopic examination of bacteria, automated digital systems represent a possible improvement in the identification and counting of microorganisms, eliminating observer bias, reducing costs and analysis time. A computational program was developed to evaluate the utility of coherent digital systems with color correlation for the recognition of Vibrio cholerae 01 in laboratory cultures and environmental samples stained with fluorescent monoclonal antibody. We tested 94 positive and 115 negative laboratory samples and, 33 positive and 34 negative environmental samples and 613 positive and 546 negative mesocosms samples. Correct identifications, cell counts and discrimination were possible in 100% of the laboratory samples. The sensitivity of the system with environmental and environmental-like samples varied between 91% to 94% and 99.5% discrimination between other bacteria or particles. Based on the absolute correlation values in red, green and blue components of the polychromatic V. cholerae 01 images (RGB channels), the algorithm to count and identify it correlated well with the peaks in the green channel output and were absent in both the red and blue channels. The discrimination criterion correlated well with the peaks present in the red and blue channels. We conclude that digital color correlation system is an useful and highly reliable tool to identify and count V. cholerae 01 in laboratory and environmental samples.