Ultrasound Lesion Diagnosis Using Basis Image Extraction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing ultrasound diagnostic systems face challenges in accurately determining lesion parts due to the risk of users selecting inappropriate ultrasound images, even when automatic detection is used, leading to reduced diagnostic accuracy.
Innovation Solution
An ultrasound diagnostic apparatus that calculates multidimensional image features, scores lesion parts based on these features, and determines the best representing image for display, using a machine learning model to enhance diagnostic accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If automatic detection of lesion parts is used to reduce user labor, then the number of images to check is reduced, but the user may not trust the detection result and select inappropriate images
Solution Approach 1:
The system provides feedback to the user by displaying the basis image that represents the determination result, allowing users to verify the automatic detection results. This feedback mechanism builds user trust while maintaining ease of operation, as users can quickly check whether the selected basis image appropriately represents the lesion characteristics.
Solution Approach 2:
The basis image acts as an intermediary between the automatic detection result and the user's final selection. Instead of directly trusting or rejecting the automatic detection, users can review the basis image as an intermediate step to confirm whether it appropriately represents the lesion, thereby resolving the trust issue while maintaining operational efficiency.
2Reliability
If multiple ultrasound images are analyzed to improve diagnostic accuracy, then the reliability of lesion determination is improved, but the complexity of image selection and processing increases
Solution Approach 1:
The system extracts a single basis image from multiple ultrasound images that best represents the determination result. This extraction process simplifies the complexity by selecting one representative image from the multiple analyzed images, maintaining diagnostic accuracy while reducing the burden of reviewing numerous images.
Solution Approach 2:
The basis image serves as a representative copy or summary of the essential characteristics from multiple ultrasound images. Instead of requiring users to examine all original images, the basis image captures and presents the key diagnostic features, reducing processing complexity while preserving diagnostic reliability.
Data Source
AI summary
There are provided an ultrasound diagnostic apparatus and a control method of the ultrasound diagnostic apparatus which allow a user to accurately perform a diagnosis for a lesion part. The ultrasound diagnostic apparatus includes an image feature calculation unit that calculates a multidimensional image feature in each of a plurality of ultrasound images in which a lesion part of a subject is imaged, by analyzing the plurality of ultrasound images; a score calculation unit that calculates a score of the lesion part in each of the plurality of ultrasound images on the basis of the multidimensional image feature; a determination unit that performs determination on the lesion part on the basis of the score calculated for each of the plurality of ultrasound images; an extraction unit that extracts the ultrasound image best representing a determination result of the determination unit from the plurality of ultrasound images, as a basis image; a monitor; and a display controller that displays the determination result of the determination unit and the basis image extracted by the extraction unit on the monitor.