CT imaging-based machine learning model: a potential modality for predicting low-risk and high-risk groups of thymoma: “Impact of surgical modality choice”

Abstract Introduction Radiomics methods are used to analyze various medical images, including computed tomography (CT), magnetic resonance, and positron emission provide information regarding the diagnosis, patient outcome, tumor phenotype, gene-protein signatures of diseases. In low-risk group, complete surgical resection is typically sufficient, whereas in high-risk thymoma, adjuvant therapy usually required. Therefore, it important distinguish between both. This study evaluated CT radiomics features thymomas discriminate low- thymoma groups. Materials total, 83 patients with were included this 2004 2019. We Radcloud platform (Huiying Medical Technology Co., Ltd.) manage imaging clinical data perform statistical analysis. The training validation datasets separated by a random method ratio 2:8 502 seeds. histopathological diagnosis was noted from pathology report. Results Four machine-learning identified differentiate group group. feature names Energy, Zone Entropy, Long Run Low Gray Level Emphasis, Large Dependence Emphasis. Conclusions results demonstrated that model multilayer perceptron classifier analysis can be on images predict thymomas. combination could useful preoperative determine approach for thymoma.