Diagnostic Accuracy of Endocytoscopy for Superficial Esophageal Neoplasia

Diagnostic Accuracy of Deep Learning-Assisted Multimodal Endocytoscopy for Superficial Esophageal Neoplasia and Precancerous Lesions: A Multicenter ，Prospective Study

Status

Enrolling by invitation

Phases

Unknown

Study type

Observational

Source

ClinicalTrials.gov

Registry ID

NCT06984055

Enrollment

600

Registered

2025-05-22

Start date

2025-03-11

Completion date

2026-12-01

Last updated

2025-05-22

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

Conditions

Esophageal Cancer, Squamous Cell

Keywords

Endocytoscopy, Esophageal Neoplasia, Precancerous Lesions, Deep Learning

Brief summary

In recent years, endocytoscopy (EC) has undergone rapid development as an advanced endoscopic technique. Characterized by real-time histological imaging and detection capabilities, EC enables visualization of living cells in the digestive mucosal layer through methylene blue staining via the endoscopic instrument channel. This technology provides ultra-magnified endocytoscopic images, allowing for identification of key cellular features including epithelial cell morphology and the shape of methylene blue-stained epithelial nuclei, thereby facilitating clear in vivo observation of microstructural characteristics in mucosal tissues - a technique commonly referred to as optical biopsy. However, current clinical application of EC in China remains limited due to insufficient practical experience and the absence of standardized diagnostic criteria for image interpretation. This study aims to comprehensively evaluate the diagnostic accuracy of domestically developed high-magnification electronic upper gastrointestinal endoscopy for esophageal carcinoma and precancerous lesions, using histopathology as the gold standard. The research objectives include establishing standardized diagnostic evaluation protocols and clarifying the clinical value of this technology through systematic validation.

Interventions

None listed

Study design

Observational model

OTHER

Time perspective

PROSPECTIVE

Eligibility

Sex/Gender

ALL

Age

18 Years to 85 Years

Healthy volunteers

Inclusion criteria

1.Patients diagnosed with superficial esophageal squamous cell carcinoma or high-grade intraepithelial neoplasia (HGIN) through endoscopic and pathological biopsy examinations, scheduled for endoscopic submucosal dissection (ESD); 2.Patients with esophageal squamous low-grade intraepithelial neoplasia (LGIN) confirmed by endoscopic and biopsy histopathological examinations within the preceding six months; 3. Patients exhibiting esophageal focal lesions detected during painless upper gastrointestinal endoscopy requiring pathological biopsy.

Exclusion criteria

1. Patients aged \<18 years or \>80 years; 2. Lesions located at the gastroesophageal junction with clinical suspicion of adenocarcinoma or precancerous lesions; 3. Patients with contraindications to sedated/anesthetized upper gastrointestinal endoscopy; 4. Patients with coagulation disorders; 5. Patients with a history of hypersensitivity to relevant medications (including anesthetics or methylene blue); 6. Patients presenting dysphagia or gastrointestinal obstructive symptoms, or those with suspected/confirmed digestive tract obstruction, stenosis, or fistulae; 7. Patients declining to provide informed consent for study participation, or any other circumstances deemed by investigators to preclude eligibility.

Design outcomes

Primary

Measure	Time frame	Description
Diagnostic accuracy	10 minutes	sensitivity and specificity

Countries

China

Outcome results