A hip joint ultrasound image analysis method, device, equipment and medium based on blind scanning
By using a blind scan method to analyze hip joint ultrasound images, and employing a cross-sectional classification branch network and a YOLOv11 network to automatically select standard coronal sections, the problem of low accuracy in hip joint development data has been solved, enabling primary care physicians to conduct rapid and accurate hip joint development examinations.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- SHENZHEN CHILDRENS HOSPITAL
- Filing Date
- 2026-01-14
- Publication Date
- 2026-06-05
AI Technical Summary
In the current technology, ultrasound image analysis methods for hip dysplasia rely on the doctor's experience, resulting in low accuracy of hip development data.
A blind scanning-based method for analyzing hip joint ultrasound images was adopted. Video streams of the hip joint were acquired using a linear array probe. Key structures were identified using a cross-sectional classification branch network and a YOLOv11 network. Ultrasound images conforming to standard coronal sections were automatically selected, and hip joint development data were extracted.
It improves the accuracy of hip development data, reduces reliance on doctors' experience, and enables primary care physicians to perform hip development examinations quickly and accurately.
Smart Images

Figure CN122156047A_ABST