Medical Image Contour Extraction for Moving Organs
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing medical image diagnostic apparatuses face challenges in accurately extracting contour points, especially for moving organs like the heart, and require significant operator input for measuring shapes, leading to insufficient accuracy and increased operational burden.
Innovation Solution
A medical image diagnostic apparatus that estimates contour positions using feature points and initial contours, extracts contours along the shape of the target part, and displays a composite image, reducing operator burden by automating the process and improving accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If multiple initial contours are generated and candidate points are designated manually, then the contour extraction can be performed, but the accuracy is insufficient when actual contour points deviate from initial contour points due to organ pulsation
Solution Approach 1:
The system automatically generates multiple initial contours and performs contour extraction without requiring manual designation of candidate points by the operator. The contour extraction unit autonomously selects actual contour points from the generated initial contours, making the system self-sufficient and eliminating manual intervention while maintaining high accuracy even when organs pulsate.
Solution Approach 2:
The system pre-generates multiple initial contours before the actual contour extraction process. By creating a set of candidate contours in advance that cover possible variations due to organ movement, the system ensures that the actual contour is included among the candidates, thereby improving extraction accuracy without requiring manual adjustment during the measurement process.
2Reliability
If multiple candidate points are input manually for each contour, then the contour extraction can be performed, but the operation burden of the examiner becomes large
Solution Approach 1:
The contour extraction unit automatically performs the entire contour extraction process without requiring manual input of candidate points for each contour. The system self-manages the selection and refinement of contour points, eliminating the time-consuming manual operation while ensuring reliable extraction results through automated algorithms that account for organ pulsation and movement.
Solution Approach 2:
The contour extraction unit is designed to handle multiple contours simultaneously using a unified automated process. Instead of requiring separate manual point designation for each contour, the system applies the same automated extraction methodology across all contours, reducing operational complexity and examination time while maintaining consistent reliability.
3Device complexity
If the contour extraction is based on fixed initial contours, then the processing is simple, but the accuracy deteriorates when organs move or pulsate
Solution Approach 1:
The system generates multiple initial contours that dynamically adapt to account for organ movement and pulsation. Instead of using a single fixed initial contour, the system creates a family of contours that cover the expected range of motion, allowing the extraction process to accurately capture the actual contour position even when organs pulsate, thereby improving measurement precision without excessive complexity.
Data Source
AI summary
A medical image diagnostic apparatus of the present invention includes: an image display unit that displays a medical image including a target part of an object; an input unit that inputs feature points of the target part; a contour position estimation unit that estimates a contour position of the target part to generate an initial contour; a contour extraction unit that extracts a contour along the shape of the target part using the feature points and the initial contour; and a control unit that displays a composite image, which is obtained by combining the extracted contour and the medical image, on the image display unit.