A Prospective Study on Deep Learning Foundation Models for Frozen Section Analysis

Status

Active, not recruiting

Phases

Early Phase 1

Study type

Observational

Source

ChiCTR

Registry ID

ChiCTR2500106350

Enrollment

Unknown

Registered

2025-07-22

Start date

2024-11-08

Completion date

Unknown

Last updated

2025-10-20

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

Conditions

tumor

Interventions

Gold Standard:The diagnosis made by multiple pathologists based on clinical information and the morphological features of frozen section slides.

Index test:Artificial intelligence based model for diagnosing tumor patients through intraoperative frozen section analysis

Eligibility

Sex/Gender

All

Age

18 Years to 87 Years

Inclusion criteria

Inclusion criteria: 1. Patients who underwent surgery and had frozen section specimens preserved 2. Age >= 18 years and <= 87 years 3. H&E pathological slides derived from intraoperative frozen section specimens

Exclusion criteria

Exclusion criteria: 1. Pathological slides exhibiting large folds or tissue detachment 2. Pathological slides with unclear scanning or inaccurate focusing 3. Patients with incomplete clinical and pathological data

Design outcomes

Primary

Measure	Time frame
Area under the curve, AUC;	—

Secondary

Measure	Time frame
Sensitivity;Specificity;False positive rate;False negative rate;	—

Countries

China

Contacts

Public ContactMuyan Cai

Sun Yat-Sen University Cancer Center

caimy@sysucc.org.cn+86 20 8734 2775

Outcome results

None listed

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