Imaging Data Analyzer ROI Correction for Statistical Reliability

Solution Overview

Problem

General statistical hypothesis tests assume a monomodality distribution, making it difficult to set reliable regions of interest (ROIs) in imaging data analysis, especially when manually set, leading to potential inaccuracies in analysis results.

Innovation Solution

An imaging data analyzer that includes a region-of-interest setting unit, data selecting unit, intensity range determination unit, and analysis execution unit to create and correct ROIs based on signal intensity distributions, ensuring a monomodality distribution for accurate analysis.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If a user manually sets an ROI in imaging data analysis, then the user can define regions of interest for analysis, but it is difficult to set the ROI completely free from micro areas that the user does not want or intend, leading to improper frequency distribution and decreased reliability of statistical hypothesis tests

Engineering Contradiction:

Improvemanual ROI settingVSAvoidstatistical hypothesis test reliability

Core Design Contradiction:

Ease of operationVSReliability

Solution Approach 1:

The system performs self-correction by automatically detecting multimodal distributions in ROI data and excluding micro areas that cause improper frequency distribution. The correction unit autonomously identifies and removes problematic regions without requiring user intervention, allowing the system to self-optimize the ROI quality while maintaining ease of manual setting.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system implements feedback by checking the frequency distribution of ROI data and automatically detecting when it deviates from the expected unimodal distribution. This feedback mechanism triggers automatic correction procedures that adjust the ROI to ensure proper distribution, thereby maintaining statistical test reliability while preserving ease of operation.

Inventive Principle:

Principle #23Feedback

2Ease of operation

If automatic ROI identification is used from optical microscopic images, then the need for user skill and experience is reduced, but micro areas that the user does not want or intend may still be included in the ROI, resulting in improper frequency distribution

Engineering Contradiction:

Improveautomatic ROI identificationVSAvoidROI boundary precision

Core Design Contradiction:

Ease of operationVSManufacturing precision

Solution Approach 1:

The system performs self-correction by automatically detecting multimodal distributions in ROI data and excluding micro areas that cause improper frequency distribution. The correction unit autonomously identifies and removes problematic regions without requiring user intervention, allowing the system to self-optimize the ROI quality while maintaining ease of automatic setting.

Inventive Principle:

Principle #25Self-service

Solution Approach 2:

The system performs preliminary verification by checking the frequency distribution of automatically identified ROI data before proceeding with statistical analysis. This preliminary check detects potential issues with ROI boundaries and triggers corrective actions, ensuring that the ROI is properly configured before analysis begins.

Inventive Principle:

Principle #10Preliminary action

3Productivity

If statistical hypothesis tests are performed on data with improper frequency distribution, then analysis can be conducted, but the reliability of the test results decreases due to violation of test assumptions

Engineering Contradiction:

Improveanalysis speedVSAvoidtest result reliability

Core Design Contradiction:

ProductivityVSReliability

Solution Approach 1:

The system performs preliminary verification by checking the frequency distribution of ROI data before proceeding with statistical analysis. This preliminary check detects potential issues with ROI boundaries and triggers corrective actions, ensuring that the ROI is properly configured before analysis begins, thus maintaining both speed and reliability.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The system performs self-correction by automatically detecting multimodal distributions in ROI data and excluding micro areas that cause improper frequency distribution. The correction unit autonomously identifies and removes problematic regions without requiring user intervention, allowing the system to self-optimize the ROI quality while maintaining ease of automatic setting.

Inventive Principle:

Principle #25Self-service

Data Source

PatentUS11651603B2Imaging data analyzer

Publication Date: 2023.05.16 SHIMADZU CORP

AI summary

When a user designates a region of interest for a plurality of groups targeted for difference analysis in a microscopic observation image of a sample, an m/z candidate search unit searches for candidates for m/z presumed to differ, based on collected mass spectral data. An intensity histogram creation unit processing unit creates and displays a graph showing a frequency distribution of peak intensities at measurement points included in the ROI of the groups for each of the m/z candidates. If this graph exhibits multimodality, the data distribution is not suitable for a statistical hypothesis test. An intensity range determination unit limits an intensity range in accordance with a user's instruction. Then, ROI correction unit corrects the ROI so as to include only measurement points with peak intensities within the limited intensity range. A test processing unit performs a statistical hypothesis test using the data corresponding to the corrected ROI.