Neural Network Biopsy Path Correction for Prostate Deformation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current mpMRI-TRUS fusion techniques for prostate cancer biopsy guidance face challenges due to misalignment of biopsy locations over time, leading to suboptimal targeting due to prostate movement and deformation between image registration and biopsy.
Innovation Solution
An ultrasound imaging system utilizing a neural network to identify and spatially delineate cancerous tissue types during a biopsy, generating a tissue distribution map and determining a corrected biopsy path based on real-time ultrasound data, incorporating user input for prioritization and feasibility constraints.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If image registration is performed once at the beginning, then the initial alignment is established, but the prostate movement and deformation between registration and biopsy causes misalignment
Solution Approach 1:
The system continuously monitors the prostate's position and deformation during the biopsy procedure and provides real-time feedback to adjust the biopsy path. The neural network analyzes ultrasound images to detect changes in tissue position and updates the target coordinates dynamically, ensuring accurate targeting despite prostate movement between initial registration and biopsy execution.
Solution Approach 2:
The biopsy path is transformed from a static pre-planned trajectory to a dynamic adjustable path that adapts to real-time prostate position changes. The system allows continuous modification of the biopsy path based on live ultrasound feedback, enabling the needle to be repositioned to compensate for prostate deformation and maintain precision throughout the procedure.
2Ease of operation
If traditional TRUS imaging is used, then real-time imaging is available, but cancerous tissue cannot be distinguished from healthy tissue
Solution Approach 1:
A neural network serves as an intermediary between the TRUS images and the biopsy targeting decision. The neural network receives real-time ultrasound images, processes them to identify cancerous tissue characteristics, and outputs corrected biopsy path recommendations. This intermediary layer enhances the diagnostic capability of TRUS by automatically detecting cancerous regions that are not visually distinguishable to human operators.
Solution Approach 2:
The system changes the parameters used for tissue characterization by applying deep learning models that analyze multiple ultrasound image features simultaneously. The neural network transforms standard TRUS images into enhanced representations that highlight cancerous tissue, effectively changing the detection parameters from simple echo intensity to multi-parameter neural network analysis.
3Measurement precision
If mpMRI-TRUS fusion is used, then cancerous tissue location is identified, but misalignment occurs due to prostate movement between registration and biopsy
Solution Approach 1:
The system incorporates continuous feedback loops that monitor prostate position and deformation during the biopsy procedure. The neural network analyzes real-time ultrasound images to detect deviations from the planned biopsy path and provides feedback to correct the targeting, ensuring reliable biopsy results even when prostate movement occurs between initial mpMRI-TRUS registration and the actual biopsy execution.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Improves diagnostic precision and treatment decisions by ensuring accurate targeting of specific cancerous tissue grades, reducing the impact of prostate movement and deformation on biopsy accuracy.
Implementation Method 1
an ultrasound transducer configured to transmit ultrasound pulses and detect echo signals responsive to the transmitted pulses
Data Source
AI summary
The present disclosure describes ultrasound imaging systems and methods configured to delineate sub-regions of bodily tissue within a target region and determine a biopsy path for sampling the tissue. Systems may include an ultrasound transducer configured to image a biopsy plane within a target region. A processor communicating with the transducer can obtain a time series of sequential data frames associated with echo signals acquired by the transducer and apply a neural network to the data frames. The neural network can determine spatial locations and identities of various tissue types in the data frames. A spatial distribution map labeling the coordinates of the tissue types identified within the target region can also be generated and displayed on a user interface. The processor can also receive user input, the neural network determines spatial locations and identities of a plurality of via the user interface, indicating a targeted biopsy sample to be collected, which can be used to determine a corrected biopsy path.