一种基于AI超声的动态乳腺密度风险评估方法
By using an AI-based ultrasound-based dynamic breast density assessment method, a spatiotemporal evolution topology map is constructed using multi-temporal ultrasound image sequences and deep convolutional neural networks. This solves the problem that traditional static assessment cannot capture dynamic changes in breast density, and enables accurate full-life-cycle prediction of breast health status.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- FUJIAN PROVINCIAL HOSPITAL
- Filing Date
- 2026-04-16
- Publication Date
- 2026-07-17
AI Technical Summary
Existing breast density assessment techniques mainly rely on static cross-sectional image analysis at a single time point, which cannot capture the dynamic changes in breast density over time. This results in insufficient sensitivity to subtle hyperplasia or degeneration trends within the breast, failing to meet the clinical need for accurate prediction of breast health status throughout the entire life cycle.
By acquiring multi-temporal ultrasound image sequences with physiological cycle timestamps, an elastic registration operation is performed to establish a unified spatial coordinate system across time dimensions. A deep convolutional neural network is used to extract spatial density feature maps and construct a spatiotemporal evolution topology map that maps the nonlinear changes in breast tissue morphology. Finally, the spatiotemporal features are synchronously aggregated through a spatiotemporal graph convolutional neural network to output a dynamic evolution trajectory vector that characterizes the hyperplasia and degeneration process inside the breast.
It achieves precise capture of the dynamic reconstruction process and nonlinear change law of breast tissue under the influence of endocrine and physiological cycles, improves the sensitivity to the evolution trajectory of potential physiological abnormalities, eliminates the lag and one-sidedness of single time point assessment, and outputs highly sensitive disease early warning status and breast health prediction throughout the life cycle.
Smart Images
Figure CN122050848B_ABST