Medical Imaging Device Sparse Scan Pattern Registration
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Medical imaging devices face challenges in performing image registration operations at a higher speed for a given image data acquisition rate, which limits their efficiency in applications like feature tracking and imaging stabilization.
Innovation Solution
A medical imaging device and method that utilize an optical imaging module to generate image data by acquiring samples indicative of optical properties and mapping them to corresponding pixels, with a control module controlling the imaging to follow a sparse scan pattern that covers a larger image area than a dense scan, allowing for improved registration efficacy and higher frame rates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a dense scan pattern is used to acquire image data, then the image coverage is complete and detailed, but the data rate increases and frame rate decreases
Solution Approach 1:
The patent segments the image acquisition process by dividing it into two distinct scan patterns: a dense scan for acquiring reference image data with high detail, and a sparse scan for acquiring real-time image data at high frame rates. This segmentation allows each scan type to be optimized for its specific purpose, resolving the contradiction between image fidelity and frame rate
Solution Approach 2:
The patent applies partial action by using a sparse scan pattern that acquires only a subset of pixels (e.g., 10-20% of total pixels) at high frame rates for real-time tracking, while using a dense scan pattern for complete reference imaging. This partial acquisition approach maintains adequate image fidelity for feature tracking while dramatically increasing frame rate
2Productivity
If a sparse scan pattern is used to increase frame rate, then the data rate decreases and processing becomes faster, but the image coverage area is reduced
Solution Approach 1:
The patent resolves the coverage area limitation by transitioning from a single-frame approach to a temporal dimension approach. Multiple sparse scans are accumulated over time to reconstruct a complete reference image, while sparse scans continue to provide high-frame-rate real-time imaging. This adds the time dimension to compensate for the reduced spatial coverage in individual frames
Solution Approach 2:
The patent performs preliminary action by first acquiring a complete reference image using a dense scan pattern, then using this pre-acquired reference for registration during subsequent sparse scanning. This preliminary reference acquisition enables the sparse scans to achieve high frame rates without sacrificing the ability to cover the complete image area through registration against the reference
3Productivity
If image registration is performed at higher speed, then feature tracking efficiency improves, but the complexity of processing increases
Solution Approach 1:
The patent extracts and processes only the essential features from the sparse image data for registration purposes, rather than processing complete high-resolution images. By extracting key features and parameters needed for registration, the system achieves high-speed processing while managing complexity, as the registration algorithm works with reduced data sets
Data Source
AI summary
A medical imaging device comprising: an optical imaging module which generates an image of a region of a body part by acquiring image samples at respective sample locations in the region and mapping the samples to corresponding image pixels; a control module which controls the optical imaging module to acquire a first sequence of samples whose corresponding pixel locations follow a path which is spanned by pixels corresponding to the first sequence of samples and extends over a greater portion of the image than an arrangement in an array of a sequence of pixels corresponding to a second sequence of samples that the optical imaging module is operable to acquire, wherein a sum of distances between adjacent pixels in the sequence of pixels is equal to a length of the path; and a registration module which registers the image against a reference image.