Medical Image Registration via Virtual Coordinate Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current medical imaging technologies face challenges in accurately and quickly registering real-time medical images, particularly due to the limitations of ultrasonic wave images in clarity and the inability of MR or CT images to reflect real-time changes caused by patient movement during treatments.
Innovation Solution
A method and apparatus for medical image registration that maps virtual coordinate systems between different medical imaging devices, allowing for the accurate tracking of anatomical features and movement by detecting similarities in anatomical objects using techniques like graph cut and Gabor wavelet analysis, and generating coordinate conversion functions to align and overlay images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Speed
If ultrasonic wave images are used for real-time imaging during medical treatment, then real-time observation capability is improved, but image clarity and sharpness deteriorate
Solution Approach 1:
The patent combines real-time ultrasonic wave images with pre-acquired high-resolution MR or CT images through image registration technology. The system merges the real-time capability of ultrasonic imaging with the high clarity of MR/CT imaging by overlaying registered anatomical information onto real-time ultrasonic images, allowing simultaneous achievement of real-time observation and high image quality.
Solution Approach 2:
The patent introduces virtual coordinate systems as an intermediary mechanism to bridge different imaging modalities. By establishing correspondence between coordinate systems of ultrasonic, MR, and CT images through anatomical feature detection and similarity measurement, the system enables accurate registration and overlay of images from different sources without requiring direct physical integration of imaging devices.
2Measurement precision
If MR or CT images are used for high clarity anatomical visualization, then image sharpness is improved, but real-time capability deteriorates
Solution Approach 1:
The patent performs preliminary action by acquiring high-resolution MR or CT images before the medical treatment procedure. These pre-acquired images serve as reference anatomical maps that are registered to the real-time ultrasonic coordinate system. This allows the system to leverage the high clarity of pre-acquired MR/CT images while maintaining real-time capability during the actual treatment through continuous ultrasonic imaging.
Solution Approach 2:
The patent creates a virtual copy of the high-resolution anatomical information from MR or CT images and overlays it onto the real-time ultrasonic image coordinate system. Through image registration that matches anatomical features between modalities, the system copies the detailed anatomical structure information from the high-quality pre-acquired images and presents it in the real-time ultrasonic framework, enabling both high clarity and real-time observation.
3Measurement precision
If manual image registration methods are used to align medical images from different sources, then registration accuracy can be achieved, but processing time and efficiency deteriorate
Solution Approach 1:
The patent implements self-service automation where the system automatically performs image registration without requiring manual intervention. The automated registration process uses computational algorithms to detect anatomical features, calculate similarity between images, and determine optimal coordinate transformations. This self-service approach eliminates time-consuming manual alignment procedures while maintaining high registration accuracy through sophisticated automated feature matching and similarity measurement techniques.
Solution Approach 2:
The patent changes the parameter of registration automation from manual to automated execution. By implementing automated algorithms that automatically detect anatomical landmarks, calculate similarity metrics, and compute coordinate transformations, the system transforms the registration process from a manual operation to an automated computational task. This parameter change in automation level significantly improves processing speed while maintaining accuracy through algorithmic optimization.
Data Source
AI summary
A method of medical image registration with respect to a volume of interest (VOI) and an apparatus for performing the method are provided. In one embodiment, the method includes obtaining a first medical image of a selected section of the VOI, from a first medical apparatus, detecting a sectional image corresponding to the selected section from second medical images previously captured of the VOI, based on an anatomical feature appearing in the first medical image, mapping virtual coordinate schemes of the first and second medical images to produce a mapped virtual coordinate scheme, based on the detected sectional image and the first medical image, and tracking a movement of a section of the VOI captured by the first medical apparatus in the second medical images by using a mapped virtual coordinate scheme.