Features of the hypointense solid lesions in the female pelvis on T2-weighted MRI

Epstein-Barr Virus-Associated Smooth-Muscle Tumor of the Brain

Epstein-Barr virus, a herpesvirus, has been associated with a variety of cancers, including Burkitt, Hodgkin, and non-Hodgkin lymphomas; posttransplant lymphoproliferative disorders; gastric carcinoma; and nasopharyngeal carcinoma, in both immunocompetent and immunocompromised individuals. Previous studies have established a connection between Epstein-Barr virus and the development of smooth-muscle tumors. Smooth-muscle tumors of the brain are very rare and are often misdiagnosed as meningiomas on imaging. To our knowledge, advanced imaging findings such as MR perfusion of smooth-muscle tumors of the brain have never been reported. We describe the radiologic and pathologic features of the Epstein-Barr virus-associated smooth-muscle tumors of the brain in a person with newly diagnosed advanced HIV.

Computed tomography and magnetic resonance imaging findings in dogs with vaginal leiomyoma and leiomyosarcoma

BACKGROUND

In humans, magnetic resonance imaging (MRI) is preferred over computed tomography (CT) for the assessment of pelvic lesions. Although CT findings of several pelvic tumours have been reported in veterinary medicine, MRI findings are limited.

OBJECTIVES

The purpose of this study was to retrospectively compare the CT and MRI findings in dogs with vaginal leiomyoma and leiomyosarcoma.

METHODS

This retrospective study of five dogs compared the CT and MRI findings of intrapelvic lesions, including vaginal leiomyoma (n = 4) and leiomyosarcoma (n = 1). No invasion of the surrounding tissue was detected on histopathological examination. In this retrospective study, the following parameters of CT and MRI were recorded for each dog: the border between the lesion and the adjacent pelvic organs, including the prostate, rectum or urethra; signal intensity (SI) of the lesion; enhancement pattern; presence of haemorrhage; necrosis or cystic areas and lymphadenopathy. Because SI on MRI is affected by cell density, tumour cell density was analysed using a microscope slide.

RESULTS

In vaginal leiomyoma, the border between the lesion and the surrounding pelvic organ tends to be clearer on MRI than on CT. In vaginal leiomyosarcoma, the border was comparable between MRI and CT. Each lesion showed heterogeneous enhancement on CT and MRI scans. In each lesion, the assessment of haemorrhage, necrosis, cystic areas and lymphadenopathy was comparable between MRI and CT. The SI of the lesion on T2WI of the vaginal leiomyoma and leiomyosarcoma were hyperintense in four cases (4/4; 100%) and mixed intense in one case (1/1; 100%), respectively. The cell density of leiomyosarcoma is higher than that of leiomyomas.

CONCLUSIONS

The SI on T2WI may be useful for differentiating leiomyoma from leiomyosarcoma. MRI may be useful to differentiate vaginal leiomyomas from leiomyosarcomas and evaluate margins.

Imaging Features of Carcinosarcoma Arising from Adenofibroma of the Uterus: A Case Report

BACKGROUND

Adenofibroma is a benign tumor composed histologically of epithelial elements and mesenchymal stroma. Carcinosarcoma is a malignant neoplasm that contains elements of carcinoma and sarcoma. Carcinosarcoma arising from adenofibroma of the uterus has never been reported in the literature in English language. Case Presentaion: We report a case of a 56-year-old woman who complained vaginal spotting persisting for several months. We described here for the first time a case of malignant transformation of uterine endometrial adenofibroma into carcinosarcoma that was depicted as a multilocular cystic lesion with enhancing solid portions and stalk-like structure between the myometrium and endometrial mass.

Machine Learning Principles for Radiology Investigators

Artificial intelligence and deep learning are areas of high interest for radiology investigators at present. However, the field of machine learning encompasses multiple statistics-based techniques useful for investigators, which may be complementary to deep learning approaches. After a refresher in basic statistical concepts, relevant considerations for machine learning practitioners are reviewed: regression, classification, decision boundaries, and bias-variance tradeoff. Regularization, ground truth, and populations are discussed along with compute and data management principles. Advanced statistical machine learning techniques including bootstrapping, bagging, boosting, decision trees, random forest, XGboost, and support vector machines are reviewed along with relevant examples from the radiology literature.

Imaging features of complex solid and multicystic ovarian lesions: proposed algorithm for differential diagnosis

Complex solid and multicystic ovarian lesions are broad-spectrum diseases, ranging from benign to malignant. This article describes the broad-spectrum and imaging features of complex solid and multicystic ovarian lesions and illustrates an algorithmic approach to such lesions, focusing on the ultrasonography and magnetic resonance imaging features that allow one to hone the differential diagnosis. Multimodality imaging workup plays an important role in the characterization and differential diagnosis of these diseases. Also, knowledge of the clinical setting and imaging features for the spectrum of complex solid and multicystic ovarian lesions can lead to appropriate management.