Multi-center Study of Deep Learning AI in Breast Mass

• Conditions: Breast Neoplasms

• Registration Number: NCT05443672
• Lead Sponsor: Cancer Institute and Hospital, Chinese Academy of Medical Sciences

Brief Summary

This multi-center study intends to evaluate the value of the detection and differential diagnosis of breast mass using deep learning AI-based real-time ultrasound examination.

Detailed Description

As the most common cancer expected to occur all over the world, extensive population screening plays a very important role in the early diagnosis and prognosis of the breast cancer. X-ray and ultrasound are the most commonly used screening methods, and ultrasound is especially important for Asian women with dense breasts. However, ultrasound is greatly affected by the operator's skill and experience, and the diagnostic accuracy varies greatly.

Artificial intelligence (AI) is a new method emerging in recent years, active in many medical fields and can effectively improve the diagnostic efficiency. However, previous researches on the application of AI in ultrasound are focused on single or multi-modality static ultrasound images. This multi-center study intends to evaluate the value of the detection and differential diagnosis of breast mass using deep learning AI-based real-time ultrasound examination.

Study Timeline

• Study Start: 2021-08-12
• Status Verified: 2022-06
• Study First Submitted: 2022-06-29
• Study First Submitted QC: 2022-06-29
• Last Update Submitted: 2022-06-29
• Study First Posted: 2022-07-05
• Last Update Posted: 2022-07-05
• Primary Completion: 2022-08-31
• Study Completion: 2023-08-31

Recruitment & Eligibility

• Status: UNKNOWN
• Sex: Female
• Target Recruitment: 1122

Inclusion Criteria

1. Females who undergo ultrasound examination for a complaint of breast lesion;
2. The breast lesion that will obtain definite pathological diagnosis or follow-up at least two years.

Exclusion Criteria

1. The breast lesion that has received CNB or FNA;
2. The breast cancer patient who has received neoadjuvant chemotherapy.

Study & Design

• Study Type: OBSERVATIONAL
• Study Design: Not specified

Primary Outcome Measures

Name	Time	Method
Diagnostic performance of breast mass using deep learning AI-based real-time ultrasound examination	12 months	Pathology as a gold standard, to evaluate the diagnostic performance (sensitivity, specificity and accuracy)

Trial Locations

Locations (1)

National Cancer Center/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College; 🇨🇳 Beijing, Beijing, China