Background: To assess the feasibility of texture analysis (TA) based on spectral attenuated inversion-recovery T2\nweighted magnetic resonance imaging (SPAIR T2W-MRI) for the classification of hepatic hemangioma (HH), hepatic\nmetastases (HM) and hepatocellular carcinoma (HCC).\nMethods: The SPAIR T2W-MRI data of 162 patients with HH (n=55), HM (n=67) and HCC (n=40) were retrospectively\nanalyzed. We used two independent cohorts for training (n = 112 patients) and validation (n = 50 patients). The TA\nwas performed and textual parameters derived from the gray level co-occurrence matrix (GLCM), gray level gradient\nco-occurrence matrix (GLGCM), gray-level run-length matrix (GLRLM), Gabor wavelet transform (GWTF), intensity-sizezone\nmatrix (ISZM), and histogram features were calculated. The capacity of each parameter to classify three types of\nsingle liver lesions was assessed using the Kruskal-Wallis test. Specificity and sensitivity for each of the studied\nparameters were derived using ROC curves. Four supervised classification algorithms were trained with the most\ninfluential textural features in the classification of tumor types. The test datasets validated the reliability of the models.\nResults: The texture analyses showed that the HH versus HM, HM versus HCC, and HH versus HCC could be\ndifferentiated by 9, 16 and 10 feature parameters, respectively. The model�s misclassification rates were 11.7, 9.6\nand 9.7% respectively. No texture feature was able to adequately distinguish among the three types of single\nliver lesions at the same time. The BP-ANN model had better predictive ability.\nConclusion: Texture features of SPAIR T2W-MRI can classify the three types of single liver lesions (HH, HM and HCC)\nand may serve as an adjunct tool for accurate diagnosis of these diseases.
Loading....