Automated Microbial Colony Counting Through Temporal Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing imaging technologies for detecting microbial growth on culture plates are difficult to automate due to their highly visual nature, making it challenging to accurately count colonies, especially when they are of different sizes and shapes, touch each other, or form confluent regions, which complicates early detection and enumeration.
Innovation Solution
An automated method for evaluating growth on plated media using digital image analysis, including threshold-based classification of colonies, matrix-assisted laser desorption ionization (MALDI) for susceptibility testing, and statistical analysis to estimate colony forming units (CFUs) based on streaking patterns and image contrast, allowing for early detection and enumeration.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If colonies are allowed to grow longer to increase contrast with background and each other, then detection accuracy improves, but colonies begin to touch and form confluent regions making counting more difficult
Solution Approach 1:
The system performs preliminary imaging of culture plates at multiple time points during incubation, capturing colony growth progression before confluence occurs. This preliminary action enables later computational analysis to distinguish individual colonies even when they eventually merge, by referencing their separate positions and growth patterns at earlier time points.
Solution Approach 2:
The invention transitions from two-dimensional spatial analysis of a single image to four-dimensional analysis by adding the time dimension. By capturing images at multiple time points during incubation, the system creates a temporal sequence that allows computational algorithms to track colony growth patterns, distinguish individual colonies, and estimate counts even when colonies form confluent regions in the final image.
2Productivity
If automated colony counting is implemented to improve productivity, then labor efficiency improves, but accuracy deteriorates when colonies are touching or in confluent regions
Solution Approach 1:
The system captures multiple preliminary images during the incubation process before final colony formation. These preliminary images serve as reference data that enable automated algorithms to accurately count colonies by tracking their growth trajectories, even when colonies eventually touch or merge in the final image, thus maintaining both automation and accuracy.
Solution Approach 2:
The system uses feedback from temporal image sequences to improve counting accuracy. By analyzing how colonies grow and change between time points, the automated system receives feedback about colony boundaries and identities, enabling it to distinguish individual colonies even when they appear merged in the final image, thereby maintaining high accuracy while preserving full automation.
3Speed
If digital image analysis is used to enable automated workflow, then speed improves, but the ability to handle highly visual inspection tasks deteriorates
Solution Approach 1:
The invention replaces the mechanical/visual inspection process with computational image analysis. Instead of relying on human visual inspection of culture plates, the system uses digital image processing algorithms to automatically detect, track, and count colonies based on temporal image sequences, thereby achieving both high speed and effective handling of visual inspection tasks through substitution with automated computational methods.
Solution Approach 2:
The system adds the time dimension to visual inspection by capturing images at multiple time points during incubation. This temporal dimension provides additional information that enables automated algorithms to distinguish colonies from artifacts, track growth patterns, and make accurate counting decisions, thereby overcoming the limitations of single-image analysis and enabling effective automated visual inspection at high speed.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables accurate and efficient automated counting of colonies, even in confluent regions, by leveraging digital image processing and statistical analysis to distinguish between colonies and artifacts, facilitating timely and cost-effective laboratory workflows.
Implementation Method 1
identifying at least one of the colonies using matrix-assisted laser desorption ionization (MALDI)
Data Source
Figure 1
Figure 2
Figure 3A~3B
AI summary
An imaging system and method for microbial growth detection, counting or identification. One colony may be contrasted in an image that is not optimal for another type of colony. The system and method provides contrast from all available material through space (spatial differences), time (differences appearing over time for a given capture condition) and color space transformation using image input information over time to assess whether microbial growth has occurred for a given sample.