Abstract
<title>Abstract</title> <p> Background This study aimed to develop a computed tomography (CT) radiomics nomogram to predict long-term survival in NSCLC patients with occult pleural dissemination (PD) after thoracic surgery. Methods Data were retrospectively collected from patients diagnosed with NSCLC and occult pleural dissemination from multiple centers. Following the exclusion of cases without follow-up data, patients were randomly assigned at a 7:3 ratio to a training set (96 cases) and a test set (40 cases). Radiomic features were extracted from primary tumor CT images and combined with independent clinical prognostic factors to construct a nomogram model for predicting overall survival (OS). The performance of the model was evaluated using the area under the receiver operating characteristic curve (AUC). Subsequently, risk stratification was performed based on the nomogram scores. Results The radiomics nomogram achieved AUC for predicting 1-, 2-, and 3-year OS ranging from 0.783 to 0.900 in the training set and from 0.728 to 0.832 in the test set, thus demonstrating superiority over the clinical model (0.796–0.835 and 0.676-0.800, respectively). Patients were then classified according to the nomogram scores into two groups: those with high risk (score ≤ 100) and those with low risk (score > 100). The high-risk group exhibited significantly inferior OS (training set: <italic>P</italic> = 0.0009; test set: <italic>P</italic> = 0.0019) in comparison to the low-risk group. Conclusions The nomogram model integrating CT radiomic features and clinical factors enables individualized prediction of long-term survival and risk stratification for NSCLC patients with occult PD. </p>