This study is a prospective cohort study aimed at developing an artificial intelligence (AI)-based predictive model for forecasting response to neoadjuvant chemotherapy in patients with locally advanced gastric cancer, using preoperative ultrasound (US), computed tomography (CT) imaging, and liquid biopsy. The study further investigates the potential biological basis underlying the proposed model.

Inclusion Criteria:

1. Be capable of understanding and voluntarily signing the written informed consent form, which must be signed prior to conducting the designated research procedures required by the study
2. Aged ≥18 years and ≤80 years at the time of signing the informed consent form (ICF)
3. Pathologically confirmed LAGC (cT2NxM0-cT4NxM0) meeting the indications for neoadjuvant chemotherapy
4. Complete gastrointestinal contrast-enhanced ultrasound and contrast-enhanced abdominal CT examinations before chemotherapy
5. Collect peripheral blood fluid samples prior to chemotherapy (for gene and protein testing)
6. After completion of standardized neoadjuvant chemotherapy, pathological specimens can be obtained postoperatively for TRG grading
7. Able to cooperate with the implementation of this protocol

Exclusion Criteria:

1. Patients with non-primary gastric cance
2. Incomplete imaging data, failed collection of body fluid samples or substandard sample qualit
3. Interrupted chemotherapy, regimen adjustment, or incomplete surgical pathological evaluation
4. Missing follow-up data, making it impossible to evaluate chemotherapy response
5. Participation in another clinical study, or presence of other factors judged by the investigator to potentially force the study subject to withdraw from the study, such as other severe illnesses requiring concurrent treatment (including psychiatric disorders), alcoholism, drug abuse, family or social factors that may affect the safety or compliance of the study subject.

patients with histologically confirmed GC at a locally advanced stage (cT2-4N0/+M0) who received NACT

The datasets utilized and analyzed in this study are not publicly available due to patient privacy requirements and ethical restriction.