Cell Image Analysis Background Brightness Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Cell image analysis methods using learned models face accuracy reduction due to uneven background brightness caused by surface tension variations in cultivation containers, leading to decreased analysis accuracy.
Innovation Solution
A cell image analysis method that generates a corrected cell image by extracting and subtracting the background brightness component, using a median filter to reduce unevenness, and employing a learned model for classification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If cell images are captured in cultivation containers with cultivation solution, then cell analysis can be performed, but uneven brightness of the background occurs due to liquid surface height variation caused by surface tension
Solution Approach 1:
The patent extracts the background component from the cell image by filtering to create a separate background image. This separates the problematic uneven brightness from the cell information, allowing the background to be corrected independently without affecting the cell data.
Solution Approach 2:
The patent changes the brightness parameter of the background by adding the difference between a reference brightness value and the actual background brightness value to each pixel. This corrects the uneven illumination while preserving the cell structure and features.
2Extent of automation
If a learned model is used to analyze cell images with uneven background brightness, then automated cell classification is achieved, but analysis accuracy is reduced
Solution Approach 1:
The patent performs background correction as a preliminary step before applying the learned model for cell classification. By correcting the uneven brightness beforehand, the learned model receives pre-processed images with uniform background, thereby maintaining high analysis accuracy while preserving automated classification.
3Productivity
If the background brightness is not corrected, then the analysis process is simple and fast, but the accuracy of normal and abnormal cell classification is reduced
Solution Approach 1:
The patent implements a self-service background correction mechanism that automatically calculates and applies brightness correction without requiring manual intervention. The system autonomously identifies background regions, computes correction values, and applies them, maintaining fast processing speed while improving classification accuracy.
Data Source
AI summary
A cell image analysis method according to this invention includes a step of acquiring a cell image (10) including a cell (90); a step of generating a background component image (14) that extracts a distribution of a brightness component of a background (91) by filtering the cell image; a step of generating a corrected cell image (11) that is corrected based on the cell image and the background component image to reduce unevenness of brightness; and a first estimation step of estimating whether the cell in the image is a normal cell or an abnormal cell by using the corrected cell image and a learned model (6) that has learned to analyze the cell.