Abstract
PURPOSE: To assess the diagnostic performance of morphological features combined with contrast-enhanced computed tomography (CECT) texture analysis in differentiating small intestinal stromal tumors (SISTs) from primary small intestinal lymphomas (PSILs). METHODS: This retrospective study included 77 patients with pathologically confirmed SISTs and 52 patients with PSILs who underwent CECT. Clinical data (age, sex, symptoms) and CT morphological features (tumor location, growth pattern, enhancement, etc.) were analyzed. Texture parameters (entropy, contrast, homogeneity, etc.) were extracted using 3D Slicer software (version 5.6.2; https://www.slicer.org/). Statistical comparisons were performed using Student's t-test or Mann-Whitney U test. Receiver operating characteristic (ROC) curve analysis was used to evaluate diagnostic efficacy. RESULTS: Compared with PSILs, SISTs exhibited significantly higher entropy (6.21 ± 0.45 vs. 5.12 ± 0.38, P < 0.001) and contrast (45.6 ± 12.3 vs. 28.7 ± 9.4, P = 0.003), but lower homogeneity (0.32 ± 0.08 vs. 0.51 ± 0.11, P = 0.002). The combined model integrating CECT morphological and texture features achieved an AUC of 0.927 (95% CI: 0.879-0.975), outperforming CECT features alone (AUC = 0.847). CONCLUSION: The integration of CECT morphological features and texture analysis enhances the differentiation of SISTs from PSILs, offering a valuable tool for improving preoperative diagnostic accuracy and guiding clinical decision-making in intestinal tumors.