3D Tissue Layer Segmentation with Motion-Corrected Feature Alignment
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current medical imaging technologies, such as OCT, suffer from poor accuracy in layer segmentation due to the continuity of layer distribution in three-dimensional space, leading to limited diagnostic information and utilization of medical images.
Innovation Solution
A method and apparatus for performing layer segmentation on a tissue structure in a medical image using feature extraction, alignment, and segmentation networks to obtain three-dimensional layer distribution, which includes determining offsets and performing feature alignment on image features to correct for tissue movement during scanning.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If two-dimensional layer segmentation is performed on each OCT image independently, then the processing complexity is reduced, but the segmentation accuracy deteriorates due to loss of three-dimensional continuity information
Solution Approach 1:
The patent transitions from two-dimensional independent segmentation to three-dimensional segmentation by incorporating the continuity of layer distribution across multiple OCT images. The system processes a stack of two-dimensional OCT images together, utilizing the third dimension (depth/scan direction) to maintain layer continuity information, thereby improving segmentation accuracy while managing processing complexity through efficient 3D convolution operations.
2Productivity
If three-dimensional segmentation is performed directly without feature alignment, then the processing speed is improved, but the segmentation accuracy deteriorates due to tissue movement during scanning
Solution Approach 1:
The patent applies feature alignment as a preliminary step before three-dimensional segmentation. By aligning the image features of multiple OCT images based on detected offsets (caused by tissue movement during scanning), the system corrects misalignments in advance. This preliminary alignment action ensures that subsequent 3D segmentation operates on corrected data, improving accuracy without significantly impacting processing speed.
3Measurement precision
If feature alignment is performed to correct tissue movement, then the segmentation accuracy is improved, but the processing time increases
Solution Approach 1:
The patent divides the processing into distinct segmentation steps: first detecting offsets between images, then aligning features based on these offsets, and finally performing three-dimensional segmentation. This segmented approach allows the system to apply alignment only where necessary (based on detected offsets) rather than processing all images uniformly, thereby reducing overall processing time while maintaining accuracy improvements from alignment.
Data Source
AI summary
A computer device performs feature extraction on two-dimensional medical images included in a three-dimensional medical image, to obtain image features corresponding to the two-dimensional medical images. The three-dimensional medical image are obtained by continuously scanning a target tissue structure. The computer device determines offsets of the two-dimensional medical images in a target direction based on the image features. The computer device performs feature alignment on the image features based on the offsets, to obtain aligned image features. The computer device performs three-dimensional segmentation on the three-dimensional medical image based on the aligned image features, to obtain three-dimensional layer distribution of the target tissue structure in the three-dimensional medical image.