AI tool to predict lymph node spread and support doctors' decisions in early (T1) stomach cancer
Official title: Development and Validation of a Multimodal Artificial Intelligence Model for Predicting Lymph Node Metastasis in T1 Gastric Cancer and Its Impact on Physician Diagnostic Performance
Run by Hebei Medical University · NCT07124754 · Data updated 2026-11-12
This study will test whether an AI that combines clinical information, pathology slides, and CT scans can predict lymph node spread in adults with T1 gastric cancer and help doctors of different experience levels make better treatment decisions.
Quick facts
| Study type | Observational |
|---|---|
| Ages | 18 Years and up |
| Sex | All |
| Enrollment | 300 (estimated) |
| Sponsor | Hebei Medical University Academic / other |
| Drugs / interventions | chemotherapy |
| Locations | 1 site (Shijiazhuang, None Selected) |
| Status | Recruiting |
| Started | January 1, 2025 |
| Main results expected | December 30, 2025 |
| Study ends | December 30, 2025 |
| Registry record updated | August 15, 2025 |
| Trial ID | NCT07124754 on ClinicalTrials.gov |
Is this trial a fit for you?
✓Likely a good fit
Adults (≥18) with histologically confirmed primary T1 (T1a or T1b) gastric adenocarcinoma who have preoperative clinical data, CT scans, and pathology slides and who underwent or will undergo radical gastrectomy with lymph node dissection and provided consent.
✕Probably not a fit
Patients with distant metastases (M1), prior other malignancies within five years, who received neoadjuvant chemo/radiotherapy, who lack complete or good-quality preoperative imaging/pathology data, or who decline consent are unlikely to benefit from this study.
A plain-English reading of the study's rules. Only the study team can confirm whether you qualify.
What this trial studies
This observational study will develop and validate a multimodal deep-learning model that integrates clinical variables, preoperative CT imaging, and histopathology slides to predict lymph node metastasis in patients with T1 gastric adenocarcinoma. Model training and internal validation will use data from adults who underwent radical gastrectomy with lymph node dissection at the Fourth Hospital of Hebei Medical University, excluding cases with poor-quality or missing imaging or prior neoadjuvant therapy. The protocol includes a physician-reader component to compare diagnostic performance with and without AI assistance across clinicians with varying levels of experience. Findings are intended to inform preoperative decision-making about whether less invasive or more extensive surgical management is appropriate for early gastric cancer.
Why it matters
Potential benefit: If successful, the AI could help identify patients who can safely avoid extensive lymph node surgery or who need more aggressive treatment, reducing unnecessary procedures and personalizing care.
How similar studies have gone: AI models using imaging or pathology data have shown promise in cancer staging and supporting physician readers, but fully integrated multimodal models specifically predicting lymph node metastasis in T1 gastric cancer remain relatively novel and not widely validated.
Who can join: the full rules
Show full inclusion / exclusion criteria
Inclusion Criteria: Age 18 years or older Histologically confirmed primary gastric adenocarcinoma Clinical stage T1 (T1a or T1b) confirmed by endoscopy and imaging Undergoing radical gastrectomy with lymph node dissection Preoperative data available: clinical variables, CT imaging, and pathology slides Written informed consent provided Exclusion Criteria: History of other malignancies within the past 5 years Received neoadjuvant chemotherapy or radiotherapy Incomplete clinical or pathological data Poor quality or missing CT or histopathology images Patients with distant metastasis (M1) at diagnosis Inability or refusal to provide informed consent
Where it's running
-
the Fourth Hospital of Hebei Medical University
Shijiazhuang, None Selected, China Recruiting
Similar studies
Other studies for the same condition or treatment. Recruiting studies are listed first.
-
AI-powered tumor response assessment after preoperative treatment in stomach and gastroesophageal junction cancer
Researchers are testing whether an AI tool called DeepTRG can reliably measure how much cancer remains in tissue removed after preoperative therapy in stomach and GEJ cancer, without needing a human to correct the AI results.
-
AI-based prediction of who will respond to pre-surgery immunotherapy in locally advanced gastric cancer
A real-world study will test whether an AI system can predict which patients with locally advanced gastric cancer will respond to neoadjuvant chemotherapy plus PD-1 inhibitors before surgery.
-
DeepComp: AI prediction of complications after stomach cancer surgery
This project will try an AI tool called DeepComp to predict which adults with gastric adenocarcinoma scheduled for curative stomach surgery are at high risk of moderate-to-severe postoperative complications using routine preoperative CT scans and clinical data.
-
Using AI to identify lymph node metastasis in gastric cancer
This study is testing a new AI tool to see if it can better identify lymph node spread in people with gastric cancer compared to regular diagnostic methods.
-
Creating a collaborative database for gastric cancer using AI technology
This study is trying to create a shared database using AI to help doctors and researchers in Hebei Province better understand and treat gastric cancer by combining patient data from different hospitals.
-
Understanding brain changes in patients with neurotoxicity from CAR T-cell therapy
This study is trying to understand how CAR T-cell therapy affects the brain in patients who develop neurotoxicity, by looking at brain scans and certain chemicals in their spinal fluid.
How to take part
- Review the eligibility criteria above with your treating physician.
- Check the official listing on ClinicalTrials.gov for the latest contact details and recruitment status.
- Contact the study coordinator or principal investigator to ask for pre-screening. Pre-screening is free and never obligates you to enroll.