Ultrasound Image Estimation Without Coordinate Conversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The variation in pixel number and resolution with scanning direction in ultrasound image data before coordinate conversion affects the accuracy of learning and estimation processing in ultrasound diagnostic systems, particularly when using mechanical 4D probes for fetal imaging.
Innovation Solution
The ultrasound diagnostic apparatus processes pre-coordinate conversion ultrasound image data using a trained model generated through machine learning, allowing for accurate estimation of the object area without needing to match image size to the direction with the highest resolution, thereby reducing processing time and maintaining accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If coordinate conversion is performed on ultrasound image data to standardize resolution, then image display quality improves, but processing time increases and resolution in high-resolution directions degrades
Solution Approach 1:
Instead of performing coordinate conversion on ultrasound image data before estimation (conventional approach), the invention inverts the process by performing estimation first on the raw ultrasound data, then applying coordinate conversion only to the estimated results. This reversal eliminates unnecessary preprocessing steps that degrade resolution and increase processing time, while maintaining accurate estimation results.
2Measurement precision
If image size is matched to the direction with the highest resolution through coordinate conversion, then display quality improves, but processing complexity and time increase
Solution Approach 1:
The invention extracts and removes the coordinate conversion step from the preprocessing stage, applying it only to the final estimation results rather than to the entire image data processing pipeline. This extraction eliminates unnecessary complexity in the estimation process while preserving the benefits of coordinate conversion for final result presentation.
3Ease of operation
If coordinate conversion is performed before estimation, then image data is standardized, but resolution variation affects learning accuracy
Solution Approach 1:
The invention performs the estimation action before the coordinate conversion action, reversing the conventional sequence. By estimating object area on raw ultrasound data first, then converting coordinates only for the estimation results, the system avoids degrading learning accuracy through premature standardization while still achieving data standardization for final presentation.
Data Source
AI summary
An ultrasound diagnostic apparatus according to an embodiment includes a processing circuitry. The processing circuitry acquires first image data that is image data obtained during the ultrasound scan executed on the object and that is image data before the coordinate conversion corresponding to the format of the ultrasound scan. The processing circuitry uses the trained model generated through the learning using the first image data obtained during the previously executed ultrasound scan and the area including the object in the first image data to estimate the area in the acquired first image data.