Objective <p>To evaluate the added value of intratumoral and peritumoral radiomic scores (RS) of iodine density (IoD) map in the preoperative prediction of lymphovascular invasion (LVI) in gastric adenocarcinoma.</p> Methods <p>A total of 226 patients, divided into groups with and without LVI, were randomly assigned to a training set (<i>n</i> = 137), an internal validation set (<i>n</i> = 59), and an external validation set (<i>n</i> = 30) that was recruited from two centers. Clinical characteristics and tumor spectral computed tomography (CT) parameters were collected and measured. Radiomic features were extracted from the IoD map within the intratumoral and peritumoral 1&#xa0;mm, 3&#xa0;mm, and 5&#xa0;mm regions. Logistic regression was used to construct RS and to identify independent risk factors for LVI. The performance of each model was evaluated through receiver operating characteristic (ROC) curve analysis.</p> Results <p>A clinical-spectral model incorporating tumor CT-T stage, Borrmann classification, cell differentiation grade, and effective atomic number (Z-Eff) as risk factors achieved an area under the curve (AUC) of 0.841 in the training set. After the addition of intratumoral and peritumoral 5&#xa0;mm RS as a risk factor to the clinical-spectral model, a nomogram was constructed, yielding a greater AUC of 0.925. The nomogram model achieved AUC of 0.872 and 0.947 in the internal and external validation sets, respectively, demonstrating superior performance compared to the clinical-spectral model.</p> Conclusion <p>Intratumoral and peritumoral 5&#xa0;mm RS based on iodine density map can provide additional value in enhancing the preoperative prediction of LVI in patients with gastric adenocarcinoma.</p>