Related Experiment Video
Updated: May 9, 2026
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
Differentiation between benign and malignant orbital tumors using deep transfer learning features and hand-crafted
Weitao Huang1, Xingjian Xu2, Xiaowei Han1
1Department of Radiology, The Quzhou Affiliated Hospital of Wenzhou Medical University, Quzhou People's Hospital, Quzhou, China.
Objective:
To evaluate the diagnostic performance of deep learning-based radiomics (DL) and hand-crafted radiomics (HCR) in differentiating benign from malignant orbital tumors.
Methods:
A retrospective analysis was performed on CT data from 145 patients (48 benign, 97 malignant) diagnosed between December 2014 and March 2024. Two radiologists independently assessed conventional CT semantic features (e.g., lesion location, margin definition, internal density homogeneity, calcification, necrosis, and enhancement pattern). Deep transfer learning (DTL) extracted DL features, while traditional methods were used to obtain HCR features. Feature fusion, selection, and modeling were performed using the Least Absolute Shrinkage and Selection Operator (LASSO). Pathological diagnosis served as the gold standard. Model performance was evaluated using receiver operating characteristic (ROC) curves. A nomogram integrating clinical data and significant semantic features was constructed for visualization. The DeLong test and decision curve analysis (DCA) assessed model effectiveness.
Results:
Multivariate analysis confirmed that homogeneous enhancement and ill-defined/infiltrative margins were independent CT features differentiating benign from malignant tumors. A total of 14 HCR and 30 DL features were extracted; 36 features were retained after fusion. The HCR, DL, fused, and nomogram models achieved AUCs of (0.859/0.816), (0.957/0.826), (0.986/0.811), and (0.975/0.837) in the training and test cohorts, respectively. The DeLong test showed no significant difference between the fused model and the nomogram in either cohort (P = 0.090 and P = 0.198), whereas differences for other model pairs were significant (P < 0.05). DCA indicated that the nomogram provided higher clinical utility.
Conclusion:
The fused model outperformed single radiomics approaches in accuracy. The nomogram, which integrates clinical data and semantic features, demonstrated superior predictive performance and may support clinical decision-making, particularly for patients who cannot undergo invasive procedures.