CNN Lesion Detection with Semantic Description
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current Computer Aided Detection (CADe) and Diagnosis (CADx) systems lack the ability to interpret and describe medical images in semantic terms, making it difficult for medical professionals to understand the diagnostic decision process and identify lesions effectively.
Innovation Solution
A convolutional neural network (CNN) is trained to detect features and map them to semantic descriptors, allowing for the automatic detection and semantic description of lesions in medical images, generating a bounding box and semantic description that can be easily understood by medical professionals.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If traditional CADe and CADx systems are used to detect lesions in medical images, then detection capability is provided, but the systems lack the ability to interpret and describe medical images in semantic terms, making it difficult for medical professionals to understand the diagnostic decision process
Solution Approach 1:
The patent combines lesion detection and semantic description generation into a single integrated CNN system. The network simultaneously performs both functions through shared feature extraction layers, eliminating the need for separate analysis tools and providing unified semantic interpretation that enhances medical professionals' understanding of diagnostic decisions.
Solution Approach 2:
The patent introduces semantic descriptors as an intermediary layer between raw image features and diagnostic conclusions. These descriptors act as interpretable intermediates that bridge the gap between complex neural network processing and human-understandable medical terminology, allowing professionals to trace diagnostic reasoning.
2Productivity
If manual lesion analysis is performed by medical professionals, then detailed semantic understanding is achieved, but diagnosis efficiency is reduced due to time-consuming manual interpretation
Solution Approach 1:
The patent enables the system to automatically generate semantic descriptions without requiring manual intervention. The CNN performs self-service by extracting features and generating interpretable descriptors autonomously, significantly reducing the time medical professionals would otherwise spend on manual lesion analysis while maintaining detailed semantic understanding.
3Measurement precision
If comprehensive feature analysis is performed on all regions in medical images, then detection accuracy is improved, but processing time and computational resources increase significantly
Solution Approach 1:
The patent applies local quality by focusing computational analysis on specific regions of interest within the medical image. The system identifies and concentrates processing resources on areas containing lesions, rather than uniformly analyzing the entire image. This approach maintains high detection accuracy for lesion regions while reducing overall processing time and computational burden.
Data Source
AI summary
An example system includes a processor to train a convolutional neural network (CNN) to detect features, and train fully connected layers of the CNN to map detected features to semantic descriptors, based on a data set including one or more lesions. The processor is to also receive a medical image to be analyzed for lesions. The processor is to further extract feature maps from the medical image using the trained CNN. The processor is also to detect a region of interest via the trained CNN and generate a bounding box around the detected region of interest. The processor is to reduce a dimension of the region of interest based on the feature maps. The processor is to generate a semantic description of the region of interest via the trained fully connected layers.