An ultrasonic ROI self-adaptive selection method and system based on image quality feedback
By combining multi-scale image processing based on image quality feedback with deep learning, the method adaptively selects the Region of Interest (ROI), solving the problem of inconsistent ROI selection in existing technologies and improving the accuracy and consistency of ultrasound image analysis.
CN122289652APending Publication Date: 2026-06-26BEIJING CHANGQING BIOMEDICAL TECHNOLOGY CO LTD
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- BEIJING CHANGQING BIOMEDICAL TECHNOLOGY CO LTD
- Filing Date
- 2026-03-30
- Publication Date
- 2026-06-26
Smart Images
Figure CN122289652A_ABST
Abstract
This invention discloses an adaptive selection method and system for ultrasound Regions of Interest (ROIs) based on image quality feedback. The method includes: amplifying potential abnormal echo regions in ultrasound images through a multi-scale image quality perturbation mechanism; automatically determining the possible quality component structures within candidate ROI sequences based on their quality feedback characteristics; adaptively grouping and filtering the ROI sequences under the constraints of these quality component structures; and finally, combining image consistency and tissue statistical distribution characteristics to obtain high-quality ROIs that meet the requirements of quantitative analysis. This invention constructs a multi-level quality screening system encompassing explicit abnormality amplification, structural component determination, and statistical consistency verification, enabling accurate and adaptive extraction of ROI regions from medical ultrasound images.
Need to check novelty before this filing date? Find Prior Art