Motion Correction for Molecular Breast Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Molecular Breast Imaging (MBI) systems face challenges in maintaining image quality due to patient organ and lesion motion during extended scanning periods, leading to potential blurring and inaccuracies in lesion size and location determination.
Innovation Solution
A method and system for motion correction in MBI images, involving the selection of a reference image, identification of a feature of interest, calculation of correction values based on the feature's location in multiple images, and alignment of non-reference images with the reference image to reduce blurring.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If reduced pressure is used to immobilize the breast between detectors, then patient comfort is improved, but image quality deteriorates due to increased motion blur
Solution Approach 1:
The system performs preliminary actions by acquiring multiple images throughout the scanning duration before final image reconstruction. These preliminary images capture the breast position at different time points, allowing subsequent motion correction through image registration and blending operations that compensate for motion blur while maintaining patient comfort with reduced pressure.
Solution Approach 2:
The system implements feedback by analyzing the acquired images to detect motion artifacts and automatically adjusting the imaging process. The image registration algorithm compares multiple images to determine relative motion, and this feedback information is used to weight and blend images appropriately, thereby correcting motion-induced quality degradation while maintaining comfortable imaging conditions.
2Reliability
If extended scanning duration is used to obtain clinically useful information, then diagnostic accuracy is improved, but image quality deteriorates due to respiratory motion
Solution Approach 1:
The scanning process is segmented into multiple discrete image acquisitions throughout the extended scanning duration. Each individual image is acquired over a short time period to minimize motion blur, and then multiple these segmented images are combined through registration and blending to achieve the diagnostic accuracy that would require longer exposure times, thereby resolving the contradiction between scanning duration and image quality.
Solution Approach 2:
The system dynamically adjusts the imaging process by continuously acquiring images throughout the extended scanning period rather than using a single static acquisition. The image registration and blending algorithms dynamically weight and combine images based on their quality and temporal information, allowing the system to maintain high diagnostic accuracy while compensating for respiratory motion that occurs during the extended scanning duration.
3Reliability
If multiple images are acquired at different orientations, then diagnostic information is improved, but image quality deteriorates due to inconsistent breast positioning
Solution Approach 1:
The system performs preliminary image acquisitions at different orientations before final image reconstruction. These preliminary images capture the breast in various positions and orientations, and subsequent image registration algorithms use these to determine relative motion and alignment. This preliminary multi-orientation acquisition allows for accurate diagnostic information while enabling computational correction of positioning inconsistencies through the registration and blending process.
Data Source
AI summary
A method for motion correcting molecular breast imaging (MBI) images includes obtaining a plurality of two-dimensional (2D) images of a breast using a MBI system, selecting a reference image from the plurality of 2D images, selecting a feature of interest in the reference image, determining a location of the feature of interest in the reference image, calculating a correction value based on a difference in the location of the feature of interest in the reference image and a location of the feature of interest in a plurality of non-reference images, and aligning the non-reference 2D images with the reference image based on the calculated correction value.