Digital Slide Image Quality Assessment via Segmented Diagnostic Scoring
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Assessing the quality and interpretability of digital slide images in pathology is challenging due to their large size and subjective examination by humans, lacking objective measures, especially since patterns for diagnosis vary between tissue types and there's no formal agreement among pathologists on criteria.
Innovation Solution
A system and method for objectively assessing digital image quality and interpretability using a rating scale based on diagnostic tasks and criteria relevant to pathologists, allowing for automation and comparison across different image capture devices, with scoring and multiple rounds of evaluation to increase accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If digital slide images are examined at full resolution to determine quality and interpretability, then measurement precision is improved, but loss of time increases due to the large size of the images
Solution Approach 1:
The patent divides the large digital slide image into multiple smaller portions or regions. Quality assessment is performed on these segmented portions rather than the entire image at full resolution, reducing examination time while maintaining sufficient measurement precision through strategic sampling of representative areas.
Solution Approach 2:
The patent creates and uses reduced-resolution copies or representations of the digital slide portions for quality assessment. These copies allow pathologists to evaluate image quality characteristics without examining the entire original image at full resolution, thereby reducing time loss while preserving assessment accuracy.
2Reliability
If objective measures are used to assess image quality, then reliability is improved, but device complexity increases due to lack of standardized criteria
Solution Approach 1:
The patent develops a universal rating scale and standardized criteria that can be applied across different image capture devices and tissue types. This multi-functional assessment framework provides objective measures that work generally for various pathology imaging scenarios, improving reliability without requiring device-specific complex systems.
Solution Approach 2:
The patent establishes standardized parameters and criteria for image quality assessment that can be consistently applied. By defining specific measurable parameters (such as resolution, contrast, noise characteristics) and standardized rating scales, the system achieves objective assessment while keeping the complexity manageable through parameter standardization rather than requiring complex adaptive systems.
3Measurement precision
If multiple rounds of scoring are performed to increase accuracy, then measurement precision is improved, but loss of time increases
Solution Approach 1:
The patent implements multiple rounds of scoring where feedback from initial assessments is used to refine and improve subsequent scoring. Pathologists or algorithms review and adjust their assessments based on previous results, gradually improving measurement precision. This feedback mechanism allows for progressive accuracy improvement while managing time through iterative refinement rather than requiring all assessments to be perfectly accurate from the start.
Data Source
AI summary
A portion of imagery data is obtained from a digital slide and a protocol of image analysis/diagnostic tasks is performed on the portion of imagery data by a pathologist or an image analysis module. The result of each task (e.g., success or no success) is recorded and a score is determined for the portion of the imagery data. Multiple portions of imagery data from the digital slide are analyzed and scored and the various scores from the multiple portions of imagery data are calculated to determine an overall score for the digital slide. Regions of the digital slide can be scored separately. Multiple rounds of scoring (by different pathologists and/or different image analysis algorithms) may be employed to increase the accuracy of the score for a digital slide or region thereof.