Skip to content

Prediction of HER2 expression status in breast cancer by artificial intelligence

A study on the prediction of HER2 expression status in breast cancer based on pre-operative ultrasound images using artificial intelligence

Status
Recruiting
Phases
Unknown
Study type
Observational
Source
ChiCTR
Registry ID
ChiCTR2500103087
Enrollment
Unknown
Registered
2025-05-23
Start date
2024-11-05
Completion date
Unknown
Last updated
2025-05-26

For informational purposes only — not medical advice. Sourced from public registries and may not reflect the latest updates. Terms

Conditions

Breast cancer

Interventions

Gold Standard:Pathological diagnosis+Immunohistochemistry(IHC)+FluorescenceInSituHybridization(FISH)
Index test:a predictive model was constructed based on the color Doppler ultrasound images of the breast of patients before treatment to non-invasively determine the expression of the HER2 gene in bre

Sponsors

Xiang'an Hospital of Xiamen University
Lead Sponsor

Eligibility

Sex/Gender
Female

Inclusion criteria

Inclusion criteria: 1. The expression of HER2 was clear in newly diagnosed women with invasive breast cancer who underwent surgical treatment (HER2(2+) had FISH detection results); 2. Women diagnosed with breast cancer by air-core needle puncture without surgical treatment, HER2(3+) or HER2(2+), and FISH test results were amplified; 3. Pathology reports are clear and searchable; 4. Ultrasound images from tertiary hospitals are available before surgery or puncture, and the images can be exported and the images are clear; 5. Pathology, immunohistochemistry, and in situ hybridization (FISH) tests are performed in qualified laboratories and by qualified and experienced physicians or technicians. 6. The time interval between ultrasound image acquisition and surgery or puncture was less than 3 weeks;

Exclusion criteria

Exclusion criteria: 1. Ipsilateral breast surgery history; 2. History of malignant tumor treatment with radiotherapy and chemotherapy; 3. Pathological findings or clinical data were missing; 4. The ultrasound image data is missing or the image quality is poor; 5. The tumor diameter is large and the image collection is incomplete;

Design outcomes

Primary

MeasureTime frame
AUC;Sensitivity;Specifcity;Accuracy;

Countries

China

Contacts

Public ContactLi Liangtao

Xiang'an Hospital of Xiamen University

LTLI7010@163.com+86 592 2889567

Outcome results

None listed

Source: ChiCTR (via WHO ICTRP) · Data processed: Feb 4, 2026