Automated Medical Image Evaluation Using Quality Thresholds
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The evaluation of medical image data records for diagnostic findings is highly manual, prone to human error, inefficient, and varies significantly between institutions, leading to challenges in speed, cost justification, and complexity in obtaining clinical results, especially in emergencies and specialized imaging tasks.
Innovation Solution
An automated method and device that use selection algorithms and image quality measures to determine suitable automated evaluation processes for medical image data records, ensuring only processes meeting minimum quality requirements are applied, leveraging artificial intelligence for image quality and evaluation quality assessments to produce standardized and reproducible clinical evaluation results.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual evaluation processes are used for medical image data records, then diagnostic accuracy can be maintained through human expertise, but evaluation time and cost increase significantly
Solution Approach 1:
The evaluation process is segmented into multiple quality assessment stages (image quality evaluation, evaluation result quality evaluation) with different threshold levels. Automated evaluation is applied to segments of the workflow where quality thresholds can be objectively measured, while manual review is reserved for cases exceeding predefined quality thresholds, thus reducing overall evaluation time while maintaining diagnostic accuracy.
Solution Approach 2:
The system changes the parameter of evaluation quality from subjective human judgment to objective quality measures (first image quality measure and second evaluation result quality measure). By establishing quantitative thresholds for these parameters, the system enables automated decision-making for routine cases, significantly reducing evaluation time while maintaining diagnostic accuracy through structured quality gates.
2Measurement precision
If highly specialized medical personnel are recruited for complex evaluation tasks, then evaluation quality improves, but recruitment and retention difficulties arise
Solution Approach 1:
The system enables self-service evaluation through automated quality assessment mechanisms. The evaluation system automatically performs quality checks on image data records and evaluation results, comparing them against predefined thresholds. This reduces dependence on highly specialized personnel for routine quality assessment tasks, making the system more adaptable to personnel availability while maintaining evaluation quality through objective automated measures.
3Measurement precision
If three-dimensional imaging and specialized clinical evaluation results are produced, then diagnostic value increases, but cost justification becomes problematic
Solution Approach 1:
The system applies partial automated evaluation action based on quality thresholds rather than evaluating all image data records with the same level of manual review. By implementing quality gates that trigger automated or manual evaluation based on measured quality parameters, the system produces high-value diagnostic results only when quality requirements are met, improving cost efficiency while maintaining diagnostic value for appropriate cases.
4Stability of the object's composition
If manual workflow processes are standardized across institutions, then reproducibility improves, but flexibility to accommodate regional requirements decreases
Solution Approach 1:
The system implements a universal automated quality assessment framework that can function across different institutions and regions. The quality evaluation mechanisms and threshold comparisons are designed to be universally applicable while allowing configuration of region-specific parameters and thresholds. This multi-functional approach enables standardized reproducible evaluation processes that can adapt to regional requirements through parameter adjustment rather than process redesign.
Data Source
AI summary
A method is for the automated evaluation of at least one image data record of a patient recorded with a medical image recording device for the preparation of diagnostic findings. In the method, at least one item of input data describing the patient and/or the recording process and/or the examination target is determined after completion of the recording of the image data record. A selection algorithm which evaluates the image data record and the input data is used for determining at least one automated evaluation process to be applied and applicable and at least one image quality measure with regard to the evaluation process is determined by evaluating the image data record. The selected automated evaluation process is only performed for an image quality measure meeting a threshold quality requirement.