Organ Classification in Tomographic Images Using Deformation Fields
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for identifying and classifying organs in 3-dimensional tomographic medical images are manual and not reusable across different images, necessitating a need for an automated solution that can accurately map and classify organs across different image sets, including changes in shape and composition over time.
Innovation Solution
A method involving the transformation of a prototype image with labeled volume elements to match the target image using deformation fields, allowing for automatic classification and segmentation of organs by transferring labels from the prototype to the target image, utilizing both water and fat image data sets for robustness and accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual identification and labeling of organs is performed, then accuracy of organ classification can be maintained, but time consumption and productivity are significantly reduced
Solution Approach 1:
A prototype image with pre-labeled organs from a reference body is created in advance. This prototype serves as a template that can be automatically transformed and applied to new target images, eliminating the need for manual labeling in each new case while maintaining classification accuracy.
Solution Approach 2:
The organ labels and anatomical structure from the prototype image are copied to the target image through image transformation. The deformation field method allows the prototype's labeled organs to be accurately mapped onto the new image, transferring classification information without manual intervention.
2Adaptability or versatility
If manual identification is performed for each new image, then adaptability to different body variations is maintained, but the process cannot be automated and productivity suffers
Solution Approach 1:
The image transformation uses a deformation field that dynamically adapts the prototype image to match the target image's anatomical variations. The transformation parameters are calculated based on the specific geometry and structure of each target image, allowing automatic adaptation to different body variations while maintaining automation.
Solution Approach 2:
The transformation process adjusts multiple parameters including translation, rotation, scaling, and non-rigid deformation to align the prototype image with the target image. These parameter changes enable the system to adapt to different body variations automatically while maintaining high productivity.
3Productivity
If image transformation with deformation fields is applied, then automatic classification productivity is improved, but the complexity of the method increases
Solution Approach 1:
The manual mechanical process of organ identification and labeling is replaced with an automated image processing system using deformation fields and transformation algorithms. This substitution eliminates manual labor while managing complexity through computational methods.
Solution Approach 2:
The deformation field acts as an intermediary between the prototype image and the target image. It mediates the transformation process by calculating the spatial mapping required to align organs between images, simplifying the overall system architecture while enabling automatic classification.
4Adaptability or versatility
If prototype image transformation is used to map organs, then reusability across different images is achieved, but precision in matching organ locations and shapes may be compromised
Solution Approach 1:
The transformation system dynamically adjusts the deformation field parameters to precisely match the target image's organ geometry. This dynamic adaptation ensures that despite using a reusable prototype, the final organ localization and shape matching achieve high precision specific to each target image.
Solution Approach 2:
Multiple transformation parameters are optimized and adjusted during the mapping process to ensure precise alignment of organs between the prototype and target images. These parameter changes include affine transformations and non-rigid deformations that preserve anatomical accuracy while enabling prototype reusability.
Data Source
AI summary
The present invention relates to a method for classification of an organ in a tomographic image. The method comprises the steps of receiving (102) a 3-dimensional anatomical tomographic target image comprising a water image data set and a fat image data set, each with a plurality of volume elements, providing (104) a prototype image comprising a 3-dimensional image data set with a plurality of volume elements, wherein a sub-set of the volume elements are given an organ label, transforming (106) the prototype image by applying a deformation field onto the volume elements of the prototype image such that each labeled volume element for a current organ is determined to be equivalent to a location for a volume element in a corresponding organ in the target image, and transferring (108) the labels of the labeled volume elements of the prototype image to corresponding volume elements of the target image.