Skip to content

From Generic to Specialized: A Subspecialty Diagnostic System Powered by Self-Supervised Learning for Cervical Histopathology

From Generic to Specialized: A Subspecialty Diagnostic System Powered by Self-Supervised Learning for Cervical Histopathology

Status
Active, not recruiting
Phases
Unknown
Study type
Observational
Source
ChiCTR
Registry ID
ChiCTR2600121560
Enrollment
Unknown
Registered
2026-04-01
Start date
2026-04-01
Completion date
Unknown
Last updated
2026-04-14

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

Conditions

Cervical squamous epithelial lesion

Interventions

Observational group:None

Sponsors

Guangdong Province Hospital for Women and Children Healthcare
Lead Sponsor

Eligibility

Sex/Gender
Female

Inclusion criteria

Inclusion criteria: The research subjects in the verification stage mainly come from clinical cases including those from Guangdong Provincial Maternal and Child Health Hospital (GWCH). The specific inclusion requirements are as follows: 1. Time span: Clinical samples collected from August 2024 to May 2025. Specimen sources: Covering common clinical specimens such as cervical Biopsy, Conization and total hysterectomy. 2. Image quality: All tissue sections were digitized by high-resolution digital section scanners (such as SQS-600), and the images were clear at 20x magnification. 3. Pathological integrity: Digital full sections (WSI) must include the complete tissue regions required for diagnosis and have clear expert-level control labels (benign, precancerous lesions or cancer subtypes).

Exclusion criteria

Exclusion criteria: 1. Image quality defects: Digital slices have obvious scanning artifacts, blurriness, severe occlusion, or severe tissue wrinkles, which affect feature extraction. 2. Abnormal specimen processing: Severe staining unevenness, excessive decalcification, or fragmented slides occur during tissue preparation, resulting in the inability to identify key morphological features. 3. Non cervical primary tumors: For the task of verifying the diagnosis of specific subtypes of the cervix, exclude all non cervical primary metastatic tumors (excluding specific pelvic pre training studies). 4. Unclear diagnostic opinion: After review by multiple experts and pathologists, consensus cannot be reached and there is controversy over the diagnosis and classification of complex cases.

Design outcomes

Primary

MeasureTime frame
Sensitivity;Specificity;Area Under the Curve;Balanced Accuracy;F1 Score;

Countries

China

Contacts

Public ContactGao Hongyi

Guangdong Province Hospital for Women and Children Healthcare

1522712335@qq.com+86 20 39151544

Outcome results

None listed

Source: ChiCTR (via WHO ICTRP) · Data processed: Apr 17, 2026