Nomogram based on computed tomography images and clinical data for distinguishing between primary intestinal lymphoma and Crohn's disease: a retrospective multicenter study.

Abstract

Differential diagnosis of primary intestinal lymphoma (PIL) and Crohn's disease (CD) is a challenge in clinical diagnosis. To investigate the validity of the nomogram based on clinical and computed tomography (CT) features to identify PIL and CD. This study retrospectively analyzed laboratory parameters, demographic characteristics, clinical manifestations, and CT imaging features of PIL and CD patients from two centers. Univariate logistic analysis was performed for each variable, and laboratory parameter model, clinical model and imaging features model were developed separately. Finally, a nomogram was established. All models were evaluated using the area under the curve (AUC), accuracy, sensitivity, specificity, and decision curve analysis (DCA). This study collected data from 121 patients (PIL = 69, CD = 52) from Center 1. Data from 43 patients (PIL = 24, CD = 19) were collected at Center 2 as an external validation cohort to validate the robustness of the model. Three models and a nomogram were developed to distinguish PIL from CD. Most models performed well from the external validation cohort. The nomogram showed the best performance with an AUC of 0.921 (95% CI: 0.838-1.000) and sensitivities, specificities, and accuracies of 0.945, 0.792, and 0.860, respectively. A nomogram combining clinical data and imaging features was constructed, which can effectively distinguish PIL from CD.