Hill-grade Knowledge Via Integrated Neural-network for Gastroscopy

NCT06040723 | Completed

• 1 other identifier: AI04
• Study Type: observational
• Target: 195
• Locations: 1 country
• Sites: 1

Brief Summary

The Hill classification, also known as the Hill grade, is a system used to classify the severity of gastroesophageal valve incompetence, specifically related to gastroesophageal reflux disease (GERD) and hiatal hernia. This study aims to compare the ability of physicians versus an AI model to asses the Hill grade during gastroscopy.

Conditions

Gastroesophageal Hernia

Keywords

Hill grade; Gastroscopy; Gastroesophageal Flap Valve; Artificial Intelligence

Outcome Measures

Primary Outcomes (1)

• Accuracy of assessments for the Hill classification for physicians and AI method.

  Binary assessment of physician vs AI correct and erroneous predictions.

  Through study completion, an average of 5 months

Secondary Outcomes (3)

• Accuracy of assessments for each Hill grade for physicians and AI method.

  Through study completion, an average of 5 months

• Precision and recall of the assessments for each Hill from endoscopists and AI method.

  Through study completion, an average of 5 months

• Distance for label assessment from gold standard label.

  Through study completion, an average of 5 months

Study Arms (1)

Experimental: Intervention arm

All patients within the study are included in the intervention arm: The Hill classification is determined by the physician and the AI method.

Device: EndoMind

Interventions

EndoMind DEVICE

The EndoMind system is equipped with an AI model for predicting the Hill grade during gastroscopy.

Experimental: Intervention arm

Eligibility Criteria

• Age: 18 Years+
• Sex: all
• Healthy Volunteers: Yes
• Age Groups: Adult (18-64), Older Adult (65+)
• Sampling Method: Non-Probability Sample

Study Population

All adult patients that have not undergone and are not undergoing a surgery for reconstruction of the flap-valve.

You may qualify if:

• Adult patients (\>18 years)
• Scheduled gastroscopy

You may not qualify if:

• Examination level
• Previous surgical interventions or altered anatomy that prevents the proper examination of the flap valve
• Flap-valve not inspected
• Data Level:
• Image during flap-valve inspection not stored
• Expert committee not resulting in a majority vote

Sponsors & Collaborators

• Wuerzburg University Hospital: lead

Study Sites (1)

Universitätsklinikum Würzburg

Würzburg, Bavaria, 97080, Germany

Location

Related Publications (1)

• Kafetzis I, Sodmann P, Herghelegiu BE, Brand M, Zoller WG, Seyfried F, Fuchs KH, Meining A, Hann A. Prospective Evaluation of Real-Time Artificial Intelligence for the Hill Classification of the Gastroesophageal Junction. United European Gastroenterol J. 2025 Mar;13(2):240-246. doi: 10.1002/ueg2.12721. Epub 2024 Dec 12.

  PMID: 39668544 DERIVED

MeSH Terms

Conditions

Hernia, Hiatal

Condition Hierarchy (Ancestors)

Hernia, Diaphragmatic; Internal Hernia; Hernia; Pathological Conditions, Anatomical; Pathological Conditions, Signs and Symptoms

Study Design

• Study Type: observational
• Observational Model: CASE ONLY
• Time Perspective: PROSPECTIVE
• Sponsor Type: OTHER
• Responsible Party: SPONSOR

Study Record Dates

• First Submitted: September 8, 2023
• First Posted: September 15, 2023
• Study Start: October 10, 2023
• Primary Completion: December 22, 2023
• Study Completion: December 22, 2023
• Last Updated: May 9, 2024

Record last verified: 2024-05

Data Sharing

• IPD Sharing: Will not share

Locations

DE (1)