Radiomics-Based AI Model for Predicting Para-Aortic Lymph Node Metastasis in Gastric Cancer Patients

NCT06947096 | Enrolling By Invitation

• 1 other identifier: GC-RAD-AI-2025-01
• Study Type: observational
• Target: 120
• Locations: 1 country
• Sites: 1

Brief Summary

This study aims to develop and validate an artificial intelligence (AI) model based on radiomics features extracted from preoperative CT images to predict para-aortic lymph node (PALN) metastasis in patients with gastric cancer. Accurately identifying PALN metastasis before surgery can help doctors make better treatment decisions, such as whether to proceed with surgery, consider chemotherapy, or use other treatment strategies. The study will prospectively enroll patients who are diagnosed with gastric cancer and scheduled for surgery. All participants will undergo routine imaging tests, and their data will be analyzed using advanced AI techniques. The results of this study may improve the precision of preoperative staging and support personalized treatment planning for gastric cancer patients.

Conditions

Gastric Cancer; Para-Aortic Lymph Node Metastasis; Lymphatic Metastasis; Preoperative Imaging Assessment; Radiomics; Artificial Intelligence

Outcome Measures

Primary Outcomes (1)

• Diagnostic Accuracy of the AI Radiomics Model for Predicting Para-Aortic Lymph Node Metastasis in Gastric Cancer

  The primary outcome is the diagnostic performance of the radiomics-based AI model in predicting para-aortic lymph node metastasis (PALNM) in patients with gastric cancer. Performance will be evaluated by calculating the area under the receiver operating characteristic curve (AUC), sensitivity, specificity, accuracy, and predictive values. The ground truth for PALNM status will be based on postoperative pathological findings or multidisciplinary consensus diagnosis. The model's predictions will be compared with actual clinical outcomes to assess its reliability and clinical utility.

  From Preoperative Imaging to Postoperative Pathological Confirmation (Approximately 4-6 Weeks per Patient)

Interventions

Radiomics-Based AI Imaging Analysis DIAGNOSTIC_TEST

This intervention involves the development and application of a radiomics-based artificial intelligence (AI) model to analyze preoperative abdominal CT images of patients with gastric cancer. The AI algorithm extracts high-dimensional imaging features from the para-aortic region to predict the presence or absence of para-aortic lymph node metastasis (PALNM). This non-invasive method aims to assist clinicians in preoperative risk stratification and treatment planning. The model will be trained and validated using manually segmented lymph node regions and correlated with postoperative pathological findings to ensure accuracy and clinical relevance.

Eligibility Criteria

• Age: 18 Years - 80 Years
• Sex: all
• Age Groups: Adult (18-64), Older Adult (65+)
• Sampling Method: Probability Sample

Study Population

The study population will consist of adult patients diagnosed with gastric adenocarcinoma who are scheduled to undergo radical gastrectomy at a tertiary care center. All participants will have preoperative contrast-enhanced CT scans and no evidence of distant metastasis. The population represents individuals at risk of para-aortic lymph node metastasis, and is intended to reflect real-world patients who may benefit from non-invasive, AI-assisted preoperative assessment tools. Participants will be enrolled consecutively to minimize selection bias.

You may qualify if:

• Adults aged 18-80 years.
• Histologically confirmed gastric adenocarcinoma.
• Planned to undergo radical gastrectomy with or without para-aortic lymph node dissection.
• Preoperative contrast-enhanced abdominal CT scan available within 3 weeks before surgery.
• No evidence of distant metastasis on imaging.
• ECOG performance status 0-2.
• Provided written informed consent.

You may not qualify if:

• History of other malignant tumors within the past 5 years.
• Received neoadjuvant chemotherapy or radiotherapy prior to CT imaging.
• Poor-quality or incomplete CT images not suitable for radiomics analysis.
• Severe comorbidities that may affect prognosis or surgical decision-making.
• Pregnancy or breastfeeding.
• Inability to provide informed consent or comply with study procedures.

Sponsors & Collaborators

• Qun Zhao: lead
• First Hospital of Shijiazhuang City: collaborator
• Baoding First Central Hospital: collaborator
• Hengshui People's Hospital: collaborator

Study Sites (1)

the Fourth Hospital of Hebei Medical University

Shijiazhuang, None Selected, 050011, China

Location

MeSH Terms

Conditions

Stomach Neoplasms; Lymphatic Metastasis

Condition Hierarchy (Ancestors)

Gastrointestinal Neoplasms; Digestive System Neoplasms; Neoplasms by Site; Neoplasms; Digestive System Diseases; Gastrointestinal Diseases; Stomach Diseases; Neoplasm Metastasis; Neoplastic Processes; Pathologic Processes; Pathological Conditions, Signs and Symptoms

Study Design

• Study Type: observational
• Observational Model: COHORT
• Time Perspective: PROSPECTIVE
• Sponsor Type: OTHER
• Responsible Party: SPONSOR INVESTIGATOR
• PI Title: Professor

Study Record Dates

• First Submitted: April 21, 2025
• First Posted: April 27, 2025
• Study Start: January 1, 2025
• Primary Completion: June 30, 2025
• Study Completion: June 30, 2025
• Last Updated: April 27, 2025

Record last verified: 2025-04

Locations

CN (1)