Background: Myxoid tumors pose diagnostic challenges for radiologists and pathologists. All myxoid tumors can\nbe differentiated from each other using fluorescent in-situ hybridization (FISH) or immunohistochemical markers,\nexcept for myxomas and myxofibrosarcomas. Myxomas and myxofibrosarcomas are rare tumors. Myxomas are\nbenign and histologically bland, whereas myxofibrosarcomas are malignant and histologically heterogenous.\nBecause of the histological heterogeneity, low grade myxofibrosarcomas may be mistaken for myxomas on core\nneedle biopsies. We evaluated the performance of T1-weighted signal intensity (T1SI), tumor volume, and radiomic\nfeatures extracted from magnetic resonance imaging (MRI) to differentiate myxomas from myxofibrosarcomas.\nMethods: The MRIs of 56 patients (29 with myxomas, 27 with myxofibrosarcomas) were analyzed. We extracted 89\nradiomic features. Random forests based classifiers using the T1SI, volume features, and radiomic features were used\nto differentiate myxomas from myxofibrosarcomas. The classifiers were validated using a leave-one-out\ncross-validation. The performances of the classifiers were then compared.\nResults: Myxomas had lower normalized T1SI than myxofibrosaromas (p = 0.006) and the AUC using the T1SI was\n0.713. However, the classification model using radiomic features had an AUC of 0.885 (accuracy = 0.839,\nsensitivity = 0.852, specificity = 0.828), and outperformed the classification models using T1SI (AUC = 0.713) and\ntumor volume (AUC = 0.838). The classification model using radiomic features was significantly better than the\nclassifier using T1SI values (p = 0.039).\nConclusions: Myxofibrosarcomas are on average higher in T1-weighted signal intensity than myxomas.\nMyxofibrosarcomas are larger and have shape differences compared to myxomas. Radiomic features performed\nbest for differentiating myxomas from myxofibrosarcomas compared to T1-weighted signal intensity and tumor\nvolume features.
Loading....