Medical Image Contour Estimation Using Feature-Point Normalization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing contour estimation techniques in medical imaging rely heavily on operator input, leading to inaccuracies due to the dependence on the accuracy of manually input points, making it difficult to estimate appropriate shapes consistently.
Innovation Solution
An image processing apparatus that utilizes a learned model to perform semiautomatic contour estimation by normalizing input images based on predetermined feature points, allowing for accurate contour estimation of targets like the right ventricle using statistical models and coordinate transformation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual contour estimation is performed by operator, then the operator can control the process, but the accuracy depends on operator skill and input precision
Solution Approach 1:
The system performs automatic contour estimation using learned models and image data, enabling the system to serve itself without requiring manual operator intervention for the actual contour extraction, thereby reducing operator burden while maintaining accuracy
Solution Approach 2:
The patent replaces the mechanical manual operation of contour drawing by operators with an automated image processing system using learned models and coordinate transformation, substituting human manual work with computational processes
2Ease of operation
If automatic contour estimation is performed using learned models, then operator burden is reduced, but accuracy depends on model quality and input point precision
Solution Approach 1:
The patent transforms the coordinate system of the input image to match the coordinate system of the learned model, changing the spatial parameters through normalization and coordinate transformation to enable accurate model application and improve estimation accuracy
Solution Approach 2:
The patent introduces coordinate transformation and normalization as intermediary processes between the input image and the learned model, creating a bridge that enables accurate matching and improves the reliability of contour estimation
3Adaptability or versatility
If coordinate transformation is applied to normalize images, then model applicability is improved, but processing complexity increases
Solution Approach 1:
The patent performs coordinate transformation and normalization of the input image before applying the learned model, preparing the data in advance to ensure compatibility and improve model applicability without increasing the complexity of the model itself
Data Source
AI summary
An image processing apparatus comprises: a model obtaining unit configured to obtain a learned model that has learned, based on a position of a predetermined feature point, a contour of a target in an image obtained by capturing the target; an image obtaining unit configured to obtain an input image; a position obtaining unit configured to obtain a position of an input point input on the input image by a user; a normalization unit configured to obtain a normalized image generated by coordinate-transforming the input image such that the position of the input point matches the position of the predetermined feature point in the learned model; and an estimation unit configured to estimate the contour of the target in the input image using the normalized image and the learned model.