Research on Early Recurrence of Locally Advanced Gastric Cancer Based on CT Radiomics Prediction

Status

Enrolling by invitation

Phases

Unknown

Study type

Observational

Source

ClinicalTrials.gov

Registry ID

NCT07683195

Acronym

LAGC

Enrollment

900

Registered

2026-07-06

Start date

2020-01-01

Completion date

2026-08-31

Last updated

2026-07-07

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

Conditions

Gastric Cancer

Keywords

Locally Advanced Gastric Cancer, Early Recurrence, CT, Radiomics

Brief summary

This study aims to develop a model for predicting postoperative recurrence in patients with LAGC using artificial intelligence (AI) technology based on preoperative computed tomography (CT) images

Interventions

None listed

Study design

Observational model

COHORT

Time perspective

RETROSPECTIVE

Eligibility

Sex/Gender

ALL

Age

18 Years to 85 Years

Healthy volunteers

Inclusion criteria

1. pathology diagnosis of LAGC (pT2NxM0-pT4NxM0); 2. radical gastrectomy with D2 lymph node dissection (\>15 lymph nodes); 3. available clinicopathological data; 4. patients underwent contrast-enhanced abdominal CT scans within 4 weeks before surgery.

Exclusion criteria

1. preoperative treatment for LAGC (radiotherapy, chemotherapy, or systemic therapy); 2. previous malignancies; 3. unsatisfactory gastric distention or inability to identify the primary tumor; 4. image artifacts.

Design outcomes

Primary

Measure	Time frame	Description
Accuracy of early recurrence models	Immediately evaluated after the early recurrence model was built	In this study, clinical data and contrast-enhanced CT imaging data of 550 patients with locally advanced gastric cancer from our hospital were collected. Machine learning and deep learning algorithms were applied to assess the early recurrence of patients within one year after surgery. The performance of the artificial intelligence model was evaluated from two dimensions: diagnostic accuracy and stability, and quantitative analysis of its performance was conducted using indicators including the area under the curve (AUC) and the precision-recall curve (PR curve).

Countries

China

Contacts

PRINCIPAL_INVESTIGATORGuangyong Zhang

Qianfoshan Hospital

Outcome results