Medical Image Annotation via Pre-trained Lesion Detection Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing technology for annotating medical images is labor-intensive and time-consuming, requiring professionally trained doctors to manually determine disease types and select lesion areas, which hinders the collection of a large number of annotated images needed for deep learning and computer-aided diagnosis.
Innovation Solution
A method and apparatus that utilize pre-trained models, such as lesion area detection and image classification models, to automatically annotate medical images by framing and splitting lesion areas, reducing the need for manual intervention and increasing efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If doctors manually annotate medical images to ensure accuracy and reliability, then annotation quality is improved, but annotation efficiency and productivity deteriorate
Solution Approach 1:
The patent introduces pre-trained models (lesion area detection model, image classification model, and lesion type classification model) as intermediary tools between the medical image and the final annotation. These models automatically perform lesion detection, classification categorization, and type identification, reducing the manual workload while maintaining annotation quality through professional medical knowledge embedded in the models.
Solution Approach 2:
The system enables self-service annotation by allowing the pre-trained models to automatically process medical images and generate annotations without requiring continuous manual intervention. The models independently complete lesion detection, classification, and type identification, freeing doctors from repetitive manual annotation tasks while preserving annotation reliability.
2Measurement precision
If doctors manually annotate medical images to ensure accuracy, then annotation precision is improved, but time consumption increases
Solution Approach 1:
The patent applies preliminary action by using pre-trained models that have been previously trained on large datasets of medical images. These models are ready to immediately perform lesion detection, classification categorization, and type identification without requiring real-time manual analysis, significantly reducing time consumption while maintaining annotation precision through the models' learned expertise.
3Quantity of substance
If large amounts of annotated medical images are collected for deep learning and computer-aided diagnosis, then data quantity for research is improved, but labor and time costs increase
Solution Approach 1:
The system enables automated self-service annotation processing that can handle large volumes of medical images without proportionally increasing manual labor. The pre-trained models automatically perform lesion detection, classification categorization, and type identification on multiple images in parallel, making it feasible to collect and process large datasets efficiently for deep learning and computer-aided diagnosis research.
Data Source
AI summary
An embodiment of the present disclosure discloses a method and apparatus for annotating a medical image. An embodiment of the method comprises: acquiring a to-be-annotated medical image; annotating classification information for the to-be-annotated medical image, wherein the classification information comprises a category of a diagnosis result and a grade of the diagnosis result corresponding to the medical image; processing the to-be-annotated medical image using a pre-trained lesion area detection model, framing a lesion area in the to-be-annotated medical image, and annotating a lesion type of the lesion area, to enable the to-be-annotated medical image to be annotated with the lesion area and the lesion type of the lesion area; and splitting the framed lesion area from the to-be-annotated medical image with the framed lesion area to form a split image of the to-be-annotated medical image, to enable the to-be-annotated medical image to be annotated with the split image.