Curved Needle Tracking in Ultrasound Imaging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current needle tracking methods in percutaneous interventions, particularly in ultrasound-based systems, face challenges due to the assumption of straight needle visibility, which is often not true, leading to inaccuracies when needles bend within soft tissues, and difficulties in longitudinal scanning due to misalignment and image quality degradation.
Innovation Solution
A dual-mode tracking approach using a weighted polynomial fitting algorithm for visible needle segments and a straight needle tracking algorithm for non-visible segments, combined with kinematics data and tissue deformation analysis, to accurately locate and track curved needles in ultrasound images.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a straight needle tracking algorithm is used, then the tracking is simple and fast, but the accuracy deteriorates when the needle bends in soft tissue
Solution Approach 1:
The needle tracking problem is segmented into two distinct cases: visible needle tracking and non-visible needle tracking. For visible segments, image-based algorithms detect needle pixels and fit curves directly. For non-visible segments, kinematic models predict needle position based on robot arm data and tissue deformation. This segmentation allows each sub-problem to be solved with appropriate complexity, improving overall accuracy without uniformly increasing system complexity.
Solution Approach 2:
The tracking system dynamically switches between different algorithms based on needle visibility conditions. When the needle is visible in ultrasound images, image-based detection is used; when obscured by tissue, kinematic prediction takes over. This dynamic adaptation resolves the contradiction by adjusting algorithm complexity according to actual tracking needs, maintaining accuracy across varying conditions.
2Reliability
If the needle visibility is low or the needle is not directly visible, then the tracking robustness should improve alternative methods, but the measurement precision deteriorates due to artifacts and tissue deformation
Solution Approach 1:
Kinematic data from the robot arm serves as an intermediary when direct needle visualization is impossible. The robot arm's position and orientation data, combined with tissue deformation models, indirectly predict needle location. This intermediary approach maintains tracking reliability when direct imaging fails, resolving the contradiction between reliability and precision by providing an alternative measurement path.
Solution Approach 2:
The system changes tracking parameters dynamically based on visibility conditions. When visibility is low, it switches from image-intensity-based parameters to kinematic parameters (robot arm position, orientation, and tissue deformation metrics). This parameter transformation allows the system to maintain measurement precision despite changes in imaging conditions, resolving the contradiction between reliability and precision.
3Measurement precision
If polynomial fitting is used for curved needle detection, then the accuracy for curved needles improves, but the computational complexity increases
Solution Approach 1:
Polynomial fitting is applied selectively only to visible needle segments in ultrasound images, not to the entire needle path. By segmenting the needle into visible and non-visible portions, the computationally intensive polynomial fitting is performed only where necessary, improving curved needle detection accuracy while limiting computational power consumption to essential regions.
4Loss of information
If the needle is scanned in the longitudinal plane, then the complete needle path can be visualized, but the image quality degrades due to misalignment
Solution Approach 1:
The system merges information from multiple sources to reconstruct the complete needle path: ultrasound images provide visual confirmation where available, robot arm kinematics provide positional data, and tissue deformation models provide contextual information. By combining these complementary data sources, the system recovers complete needle path information without relying solely on degraded longitudinal plane imaging.
Data Source
AI summary
Provided is a system, method, and computer program product for tracking a needle. The method includes determining a visibility of the needle being inserted into a subject in an image of a sequence of images, in response to determining that the visibility satisfies a visibility threshold, detecting a location of the needle based on at least one first algorithm and a detected curvature of the needle, in response to determining that the visibility does not satisfy the visibility threshold, detecting the location of the needle being inserted based on at least one second algorithm, and tracking the location of the needle in the sequence of images based on locations detected with the at least one first algorithm and the at least one second algorithm.