Related Experiment Video
Updated: Jul 6, 2025
07:21
Automated, High-Throughput Detection of Bacterial Adherence to Host Cells
Published on: September 17, 2021
3.5K
Line image sensor-based colony fingerprinting system for rapid pathogenic bacteria identification
Hikaru Tago1, Yoshiaki Maeda2, Yusuke Tanaka1
1Division of Biotechnology and Life Science, Institute of Engineering, Tokyo University of Agriculture and Technology, 2-24-16 Naka-cho, Koganei, Tokyo, 184-8588, Japan.
Biosensors & Bioelectronics
|January 10, 2024
Summary
A new line image sensor system rapidly identifies pathogenic bacteria using colony fingerprints, achieving 96% accuracy in 10 hours. This advances microbiological testing for industrial applications.
Area of Science:
- Microbiology
- Biotechnology
- Imaging Technology
Background:
- Rapid identification of pathogenic bacteria is essential for industrial quality control, particularly in food and beverage manufacturing.
- Current bacterial identification methods face limitations in speed and throughput, hindering efficient inspection processes.
- Bacterial microcolony image-based classification offers a potential solution but is constrained by conventional imaging system's slow speed and limited field of view.
Purpose of the Study:
- To develop a novel imaging system utilizing a line image sensor for rapid, wide-field bacterial microcolony imaging.
- To establish a colony fingerprinting method for machine learning-based bacterial identification.
- To demonstrate the system's efficacy in identifying specific foodborne pathogens like Staphylococcus aureus.
Main Methods:
- Development of an imaging system employing a line image sensor capable of capturing a full Petri dish image within 22 seconds.
- Acquisition of bacterial microcolony images and extraction of discrimination parameters to create 'colony fingerprints'.
- Training a machine learning model on a dataset of colony fingerprints from 15 bacterial species for identification.
Main Results:
- The developed system successfully acquired bacterial microcolony images and generated colony fingerprints.
- The system achieved 96% accuracy in identifying Staphylococcus aureus within 10 hours of incubation.
- This performance significantly outperforms conventional mass spectrometry methods, which require 24 hours.
Conclusions:
- Line image sensor technology enables high-speed, scalable microbiological testing, overcoming limitations of conventional methods.
- Colony fingerprinting combined with machine learning offers a rapid and accurate approach for bacterial identification in industrial settings.
- This innovation streamlines microbiological analysis, reducing the need for specialized expertise and accelerating quality control processes.
Keywords:
Colony fingerprintingEarly detectionHigh-throughput testingLine image sensorMachine learningPathogenic bacteria
