Dynamic Threshold Adjustment for Medical Image Regions of Interest
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing computer-aided diagnosis systems for medical images face challenges in balancing detection rate and false positive rate, limiting their versatility across different scenarios and case characteristics due to fixed parameter thresholds.
Innovation Solution
A method and system that allow users to adjust attribute parameter thresholds in real-time, enabling the display of regions of interest in medical images based on user-input thresholds, which compares attribute parameters like confidence level, category, and size to optimize diagnostic accuracy and reading time.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If fixed parameter thresholds are used in computer-aided diagnosis systems, then the system operation is simplified, but the adaptability to different usage scenarios and case characteristics deteriorates
Solution Approach 1:
The patent implements dynamic threshold adjustment by allowing users to modify parameter thresholds in real-time during system operation. The threshold adjustment module enables doctors to change detection parameters based on different usage scenarios and case characteristics, transforming the static threshold system into a dynamic one that adapts to varying diagnostic needs.
Solution Approach 2:
The system allows changing of parameter thresholds (such as confidence level thresholds) to adapt to different diagnostic scenarios. By modifying the threshold parameters, the system can balance between detection rate and false positive rate according to specific case requirements, thereby improving versatility without complicating the core operation.
2Device complexity
If parameter thresholds are fixed, then the system complexity is reduced, but the versatility across different usage scenarios deteriorates
Solution Approach 1:
The patent introduces a dynamic threshold adjustment mechanism that allows the system to adapt to different scenarios without requiring multiple separate systems or complex reconfiguration. The threshold adjustment module provides a simple interface for modifying parameters, maintaining low operational complexity while significantly improving versatility.
Solution Approach 2:
The system achieves multi-functionality by enabling a single computer-aided diagnosis system to handle various usage scenarios through adjustable parameters. The same core detection algorithm can be adapted to different diagnostic needs by changing thresholds, making the system universal across multiple applications without increasing fundamental complexity.
3Ease of operation
If a single parameter threshold is used, then the system operation is simplified, but the ability to balance detection rate and false positive rate across different cases deteriorates
Solution Approach 1:
The patent enables changing of parameter thresholds to optimize the balance between detection rate and false positive rate. Doctors can adjust the confidence level threshold according to case characteristics - lowering it to increase detection rate when sensitivity is needed, or raising it to reduce false positives when specificity is prioritized.
Solution Approach 2:
The system transitions from static to dynamic threshold setting, allowing real-time adjustment of parameters during diagnostic workflows. This dynamic capability enables the system to adapt to different reliability requirements while maintaining simple operation through an intuitive adjustment interface.
Data Source
AI summary
Method and system for displaying one or more regions of interest of an original image. For example, a computer-implemented method for displaying one or more regions of interest of an original image includes: obtaining one or more detection results of one or more first regions of interest, each detection result of the one or more detection results corresponding to one first region of interest of the one or more first regions of interest, each detection result including image information and one or more attribute parameters for their corresponding first region of interest; and obtaining one or more attribute parameter thresholds provided by a user in real time, each attribute parameter threshold of the one or more attribute parameter thresholds corresponding to one attribute parameter of the one or more attribute parameters.


