Automatic Counting and Classification of Bacterial Colonies Using Hyperspectral Imaging

Abstract

Detection and counting of bacterial colonies on agar plates is a routine microbiology practice to get a rough estimate of the number of viable cells in a sample. There have been a variety of different automatic colony counting systems and software algorithms mainly based on color or gray-scale pictures although manual counting is still common. In microbiology, identification of presumptive-positive colonies on agar plates is predominantly done manually, which is laborious and time-consuming. This paper addresses a problem related to automatic colony segmentation and classification that can count the number of colonies according to their types. Hyperspectral imaging was used to develop a colony segmentation algorithm for detecting non-O517 Shiga-toxin producing Escherichia coli (STEC) pathogens on Rainbow agar. Hyperspectral absorbance image analysis in the visible and near-infrared spectral range from 400 to 1000 nm showed that colony morphology including size and texture was dependent on wavelength. The non-O157 STEC colonies showed dome-like absorbance profiles with local absorbance maxima. Touching colonies, causing problems for accurate counting and identification, were separated by optimally tessellating the mesh structure of local maximal points. The 428 nm was determined as the optimal wavelength for non-O157 STEC colony segmentation. The accuracy of the colony segmentation and counting algorithm was over 99 %. The average of the colony classification algorithm using automated colony segments was 92.5 %.