Automatic identification of algae: neural network analysis of flow cytometric data

Abstract

The performance of an artificial neural network for automatic identification of phytoplankton was investigated with data from algal laboratory cultures, analysed on the Optical Plankton Analyser (OPA), a flow cytometer especially developed for the analysis of phytoplankton. Data from monocultures of eight algal species were used to train a neural network. The performance of the trained network was tested with OPA data from mixtures of laboratory cultures. The network could distinguish Cyanobacteria from other algae with 99% accuracy. The identification of species was performed with less accuracy, but was generally >90%. This indicates that a neural network under supervised learning can be used for automatic identification of species in relatively complex mixtures. Incorporation of such a system may also increase the operational size range of a flow cytometer. The combination of the OPA and neural network data analysis offers the elements to build an operational automatic algal identification system.