Diagnostic Performance of 2D and 3D T2WI-Based Radiomics Features With Machine Learning Algorithms to Distinguish Solid Solitary Pulmonary Lesion

Objective: To evaluate the performance of 2D and 3D radiomics features with different machine learning approaches to classify SPLs based on magnetic resonance T2 weighted imaging (T2WI). Material and Methods: A total of 132 patients with pathologically confirmed SPLs were examined and randomly divided into training (n = 92) and testing datasets (n = 40). A total of 1692 3D and 1231 2D radiomics features per patient were extracted. Both radiomics features and basic clinical data were evaluated. A total of 1260 classification models, comprising 3 normalization methods, 2 dimension reduction algorithms, 3 feature selection methods, and 10 classifiers with 7 different feature numbers (confined to 3–9), were compared. Results: The 3D features were significantly superior to 2D features, showing much more machine learning combinations with higher area under the curve (AUC). The best performance of 3D radiomics features in the test dataset (AUC = 0.824) was higher than that of 2D features(AUC=0.740). The joint 3D and 2D features (AUC=0.812) showed similar results as 3D features. The best 3D radiomics model for the classification was a combination of principal component analysis + analysis of variance + Gaussian process. Incorporating clinical features with 3D and 2D radiomics features slightly improved the AUC to 0.836 and 0.779, respectively. Conclusions: The radiomics model based on T2WI could differentiate malignant and benign SPLs. The 3D radiomics features were better than 2D features in differentiating SPLs. The choice of classification method could lead to performance variation.