Multi-view Mammogram Analysis Reducing False Positives
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing mammography techniques primarily analyze single images, leading to higher false positive rates and inadequate detection of lesions in breast cancer screening.
Innovation Solution
A multi-view mammogram analysis method utilizing a symptom identification model to generate heat maps and a false positive filtering model to determine abnormal probabilities, with a threshold-based system for detecting lesion positions in multiple mammogram images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If single-image classification and detection is used in mammography, then the analysis process is simple, but the false positive rate increases and detection accuracy decreases
Solution Approach 1:
The patent divides the mammography analysis into multiple independent modules: a symptom identification model for initial detection, a false positive filtering model for error reduction, and a lesion position analyzing unit for precise localization. Each module processes specific aspects of the images independently, allowing the system to maintain simplicity while improving accuracy through specialized function separation.
Solution Approach 2:
The patent transitions from single-image analysis to multi-view mammogram analysis by processing mammograms from different angles and perspectives. This dimensional expansion allows the system to cross-validate findings across multiple views, reducing false positives while maintaining manageable complexity through systematic processing of each view.
2Reliability
If multi-view mammogram analysis is implemented, then false positive rate decreases and detection accuracy improves, but the analysis complexity increases
Solution Approach 1:
The patent employs a multi-functional processing system where the symptom identification model, false positive filtering model, and lesion position analyzing unit work together across multiple mammogram views. Each component serves universal purposes: detecting abnormalities, filtering false positives, and locating lesions, regardless of which specific view is being processed, thereby managing complexity through standardized multi-functional operations.
Solution Approach 2:
The patent implements preliminary symptom identification and false positive filtering before final lesion position determination. By performing these preparatory analyses on multiple views first, the system reduces the complexity of the final detection stage, as the subsequent lesion localization works with pre-processed, validated data rather than raw multi-view inputs.
Data Source
AI summary
A multi-view mammogram image analysis method, multi-view mammogram image analysis system and non-transitory computer-readable medium are provided in this disclosure. The multi-view mammogram image analysis method includes the following operations: inputting a plurality of mammogram images; utilizing a symptom identification model to determine whether the mammogram images have an abnormal state, and generating a plurality of heat maps corresponding to the mammogram images; utilizing a false positive filtering model to determine whether the heat maps have a false positive feature, and generating an abnormal probability corresponding to the heat maps; and utilizing a first threshold to determine the abnormal probability, if the abnormal probability is greater than the first threshold, detecting and outputting a lesion position corresponding to the heat maps.