NCT06646679

Intelligent Segmentation Algorithm of Ultrasonic Image

• Status: Recruiting
• Phase: —
• Study type: Observational
• Enrollment: 5,000 (estimated)
• Sponsor: Tongji Hospital · Academic / Other
• Sex: All
• Age: —
• Healthy volunteers: Accepted

Summary

This project aims to enhance the performance of ultrasonic image analysis by optimizing and refining neural network algorithms, while also collecting and constructing extensive datasets relevant to ultrasonic imagery. The algorithm will be trained and evaluated in a data-driven manner, with test results facilitating accurate segmentation of regional block images and identification of characteristic ultrasonic anatomy. This approach will significantly advance the study and development of ultrasonic technology.

Detailed description

1. Algorithm development: Develop high-performance intelligent ultrasonic image recognition and segmentation algorithms; 2. Data set: Establish ultrasound image data set, which is convenient for researchers to develop and verify algorithms; 3. Apply for patents, publish academic papers, promote and popularize technology.

Conditions

• Segmentation

Interventions

Type	Name	Description
OTHER	Ultrasonic imaging	Ultrasonic imaging

Timeline

• Start date: 2024-10-14
• Primary completion: 2026-11-02
• Completion: 2026-11-14
• First posted: 2024-10-17
• Last updated: 2024-11-05

Locations

1 site across 1 country: China