A Multi-center Study of Breast Mass Screening and Diagnosis Using Deep Learning AI-based on Real-time Ultrasound Examination
Overview
- Phase
- Not Applicable
- Intervention
- Not specified
- Conditions
- Breast Neoplasms
- Sponsor
- Cancer Institute and Hospital, Chinese Academy of Medical Sciences
- Enrollment
- 1122
- Locations
- 1
- Primary Endpoint
- Diagnostic performance of breast mass using deep learning AI-based real-time ultrasound examination
- Last Updated
- 3 years ago
Overview
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.
Investigators
Eligibility Criteria
Inclusion Criteria
- •Females who undergo ultrasound examination for a complaint of breast lesion;
- •The breast lesion that will obtain definite pathological diagnosis or follow-up at least two years.
Exclusion Criteria
- •The breast lesion that has received CNB or FNA;
- •The breast cancer patient who has received neoadjuvant chemotherapy.
Outcomes
Primary Outcomes
Diagnostic performance of breast mass using deep learning AI-based real-time ultrasound examination
Time Frame: 12 months
Pathology as a gold standard, to evaluate the diagnostic performance (sensitivity, specificity and accuracy)