The software “Colony Counting System for Easy Plate” can automatically count and analyze the colonies in imaged Easy Plate. A solution to save time and eliminate human errors using the AI technology. The poster session was presented during IAFP 2023.
Poster (PDF) is also available for download.
Kentaro Takenaka1, Shinichiro Sugiura2
(1) Kikkoman Corporation, (2) Kikkoman Biochemifa Company
Abstract
Introduction
Ready-to-use (RTU) culture media enhances food safety and productivity because of its quickness, compactness, simplicity, and visibility of colonies compared to conventional agar media. Since colony counting requires skilled technicians and is time-consuming, automated colony counting systems are used for better efficiency.
Purpose
To confirm the rapidity and accuracy of the novel “Colony Counting System for Easy Plate (CCS)” - a plate reading technology based on a general scanner with auto document feeder and analysis software.
Methods
The CCS consists of a scanner (Brother Industry, LTD., ADS-4300N) and counting software based on object detection using machine learning. Ten food samples were diluted and inoculated onto RTU-culture media, Easy Plate AC (Kikkoman Biochemifa Company) for aerobic bacteria count (AC) and Easy Plate CC for coliform bacteria count (CC) at three contamination levels and incubated according to the respective instructions for use. Differences of bacterial counts and measurement time between conventional manual methods and CCS method were evaluated.
Results
The CCS method took 5.0 to 6.4 seconds/sheet regardless of contamination level. On the other hand, conventional manual methods took 13.8 to 44.5 seconds/sheet at low contamination levels ranging from 10 to 50 CFU/sheet, the level at which manual counting methods were shown to be most rapid. The correlation coefficients for bacterial counts between manual counting and the CCS method were greater than 0.98 for both AC and CC. These results indicated that 77% of analysis time could be saved by the use of the CCS method with counting accuracy comparable to manual methods.
Significance
It was demonstrated that this system provides automated colony counting with rapidity and accuracy. An automated system can also be expected to reduce human error. Therefore, the combination of RTU-culture media and an automated system can save time and reduce costs of microbiological testing while maintaining accuracy.
Download poster (PDF) from here.
