Radiographics : a review publication of the Radiological Society of North America, Inc
-
The Liver Imaging Reporting and Data System (LI-RADS) is a reporting system created for the standardized interpretation of liver imaging findings in patients who are at risk for hepatocellular carcinoma (HCC). This system was developed with the cooperative and ongoing efforts of an American College of Radiology-supported committee of diagnostic radiologists with expertise in liver imaging and valuable input from hepatobiliary surgeons, hepatologists, hepatopathologists, and interventional radiologists. In this article, the 2017 version of LI-RADS for computed tomography and magnetic resonance imaging is reviewed. Specific topics include the appropriate population for application of LI-RADS; technical recommendations for image optimization, including definitions of dynamic enhancement phases; diagnostic and treatment response categories; definitions of major and ancillary imaging features; criteria for distinguishing definite HCC from a malignancy that might be non-HCC; management options following LI-RADS categorization; and reporting. ©RSNA, 2017.
-
In response to the recommendation of the U. S. ⋯ In this article, a collection of 15 LCS-related scenarios are presented that address situations in which the Lung-RADS guidelines are unclear or situations that are not currently addressed in the Lung-RADS guidelines. For these 15 scenarios, the authors of this article provide the reader with recommendations that are based on their collective experiences, with the hope that future versions of Lung-RADS will provide additional guidance, particularly as more data from widespread LCS are collected and analyzed. ©RSNA, 2017.
-
Deep learning is a class of machine learning methods that are gaining success and attracting interest in many domains, including computer vision, speech recognition, natural language processing, and playing games. Deep learning methods produce a mapping from raw inputs to desired outputs (eg, image classes). Unlike traditional machine learning methods, which require hand-engineered feature extraction from inputs, deep learning methods learn these features directly from data. ⋯ Recently, several clinical applications of CNNs have been proposed and studied in radiology for classification, detection, and segmentation tasks. This article reviews the key concepts of deep learning for clinical radiologists, discusses technical requirements, describes emerging applications in clinical radiology, and outlines limitations and future directions in this field. Radiologists should become familiar with the principles and potential applications of deep learning in medical imaging. ©RSNA, 2017.
-
Review
2016 Updates to the WHO Brain Tumor Classification System: What the Radiologist Needs to Know.
Radiologists play a key role in brain tumor diagnosis and management and must stay abreast of developments in the field to advance patient care and communicate with other health care providers. In 2016, the World Health Organization (WHO) released an update to its brain tumor classification system that included numerous significant changes. Several previously recognized brain tumor diagnoses, such as oligoastrocytoma, primitive neuroectodermal tumor, and gliomatosis cerebri, were redefined or eliminated altogether. ⋯ The increased emphasis on genetic factors in brain tumor diagnosis has important implications for radiology, as we now have tools that allow us to evaluate some of these alterations directly, such as the identification of 2-hydroxyglutarate within infiltrating gliomas harboring mutations in the genes for the isocitrate dehydrogenases. For other tumors, such as medulloblastoma, imaging can demonstrate characteristic patterns that correlate with particular disease subtypes. The purpose of this article is to review the changes to the WHO brain tumor classification system that are most pertinent to radiologists. ©RSNA, 2017.