Automatic Analyzer Abnormality Detection via Dynamic Criterion Generation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing automatic analyzers face challenges in setting an optimal determination criterion for detecting abnormalities in reaction processes due to variations in machine differences and sample data deviations, leading to inaccurate feature quantity representation and difficulty in setting appropriate criteria.
Innovation Solution
An abnormality determining method that generates a determination criterion based on feature quantities of measured data, using a processor to acquire and process data from a spectral detector, calculate feature quantities, and compare them against a generated criterion to determine the presence of abnormalities, incorporating acquisition conditions such as measured value range divisions and data segment information.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a determination criterion is set in advance with large tolerance to account for machine differences, then the setting is more universally applicable, but the abnormality detection accuracy deteriorates
Solution Approach 1:
The patent changes the parameters of the determination criterion dynamically based on the actual measured data. Instead of using fixed predetermined values, the system calculates feature quantities from actual measurement data and sets determination criteria (such as upper and lower limits) based on statistical parameters like mean and standard deviation of the actual data distribution. This allows the criterion to adapt to specific machine characteristics while maintaining statistical rigor for abnormality detection.
2Measurement precision
If sample data is acquired to set the determination criterion for each analyzer, then the criterion can be customized, but deviations in measured values occur making it difficult to correctly grasp the feature quantity distribution
Solution Approach 1:
The patent performs preliminary actions by accumulating a large amount of measurement data before setting the determination criterion. The system collects feature quantities from multiple measurement results under normal operating conditions, calculates statistical parameters (mean, standard deviation) from this accumulated data, and establishes the determination criterion based on these pre-calculated parameters. This preliminary data accumulation ensures reliable representation of the normal feature quantity distribution.
Solution Approach 2:
The patent segments the measurement data into multiple divided segments and calculates feature quantities for each segment separately. By dividing the data range into multiple segments and analyzing the distribution characteristics in each segment, the system can more accurately capture the overall feature quantity distribution without being affected by deviations in individual measurements. This segmentation approach improves the reliability of the determined criterion.
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
This method allows for a more appropriate determination criterion to be set for detecting abnormalities, improving the reliability of abnormality detection and data accuracy in automatic analyzers.
Implementation Method 1
an absorbance and a luminescence amount of the reaction solution are measured
Implementation Method 2
an absorbance and a luminescence amount of the reaction solution are measured
Data Source
AI summary
The present invention relates to an improved abnormality determining method for an automatic analyzer and related automatic analyzer that detects abnormalities for a reaction process of a reaction solution. The method for determining the presence of absence of abnormalities in the reaction process and performs the determination by acquiring acquisition conditions for multiple items of measured data, acquiring the measured data detected by a spectral detector that match the acquisition conditions, calculating a feature quantity of the measured data, generating a determination criterion based on the feature quantity, and determining the presence or absence of abnormalities by comparing the feature quantity of the measured data subject to the determination with the determination criterion. The acquisition condition includes information relating to a number of a divisions of a measured value range and information relating to the number of items of data in each divided segment.