Cell Image Processor Segmentation for Noise Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional cell image analyzers fail to effectively extract and analyze cells with specific characteristics, as they are often misidentified as noise and buried in data, making it difficult to distinguish meaningful phenomena from noise or contaminants.
Innovation Solution
A cell image processor and method that measures and processes characteristic quantities of cells to form distributions, allowing for the extraction of cells with continuously distributed characteristics, thereby separating specific cells from noise and enabling their analysis.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If cells with specific characteristics are eliminated as noise to improve measurement accuracy, then analytical precision is improved, but meaningful specific phenomena are lost
Solution Approach 1:
The patent segments the cell population into different groups based on characteristic quantity distributions. By dividing cells into those with continuous distributions (normal variation) and those with discontinuous distributions (specific phenomena), the system can preserve meaningful specific cells while eliminating noise, thus resolving the contradiction between measurement precision and information loss
Solution Approach 2:
Instead of eliminating all cells that deviate from the average (excessive action), the patent applies partial action by selectively eliminating only those cells with discontinuous characteristic quantity distributions. This partial approach preserves cells with meaningful specific phenomena while still eliminating noise, balancing precision improvement with information retention
2Loss of information
If all cells are processed individually to preserve specific phenomena, then information loss is reduced, but processing time increases
Solution Approach 1:
The patent segments cells into different categories based on their characteristic quantity distributions. By identifying and separating cells with discontinuous distributions (specific phenomena) from those with continuous distributions (normal variation), the system can apply different processing strategies to each segment, preserving specific information while optimizing processing efficiency
Solution Approach 2:
The patent changes the parameter used for cell selection from individual characteristic values to the distribution pattern of characteristic quantities. By analyzing whether characteristic quantities show continuous or discontinuous distributions, the system can efficiently identify specific cells without processing every cell individually in detail, thus reducing processing time while preserving information
3Productivity
If noise elimination is performed using average-based methods, then processing speed is improved, but specific cells are misidentified as noise
Solution Approach 1:
The patent changes the selection criterion from absolute deviation from average to the distribution pattern of characteristic quantities. By examining whether characteristic quantities show continuous or discontinuous distributions, the system maintains fast processing speed while dramatically improving the accuracy of specific cell identification, resolving the contradiction between productivity and measurement precision
Data Source
AI summary
Cells showing specific characters can be separated from noise, extracted, and analyzed, through processing of a cell image. It is intended to provide a cell image processor (1) comprising: a characteristic quantity measuring section (11) which processes a cell image obtained by photographing cells to measure characteristic quantities of respective cells in the cell image; a characteristic quantity distribution forming section (12) which forms the distribution of the thus measured characteristic quantities; a group forming section (13) which divides cells having continuously distributed characteristic quantities into groups in the thus formed characteristic quantity distribution; and a specific cell extracting section (14) which extracts cells having characteristic quantities falling within a predetermined range at both ends of a characteristic quantity distribution in each group, as specific cells.