The following are highlights from the current issues of RSNA’s peer-reviewed journals.
Breast MRI: State of the Art
MRI of the breast has the highest sensitivity for breast cancer detection among current clinical imaging modalities and is indispensable for breast imaging practice. While the basis of breast MRI consists of T1-weighted contrast-enhanced imaging, T2-weighted, ultrafast and diffusion-weighted imaging may be used to improve lesion characterization. Such multiparametric assessment of breast lesions allows for excellent discrimination between benign and malignant breast lesions.
In an article published online in Radiology, Ritse M. Mann, MD, PhD, Radboud University Medical Centre, Nijmegen, the Netherlands, and colleagues reviewed the current state of the art in breast MRI, with a focus on the major indications and the potential indication-based adaptations to the imaging protocol to maximize its value.
“Technical developments have improved the quality of breast MRI, allowing the acquisition of isotropic high-resolution images. In addition, multiparametric breast MRI has largely replaced the conventional approach, which was primarily based on conventional contrast-enhanced sequences alone for lesion classification. Multiparametric evaluation of lesions allows for a positive predictive value of biopsy that is comparable to that for mammography. Other techniques currently being investigated may enable a more quantitative approach to MRI evaluation, further improving reproducibility and consistency over clinical sites,” the authors conclude.
To read the full article, go to RSNA.org/Radiology.
Automated Organ-Level Classification of Free-Text Pathology Reports to Support a Radiology Follow-Up Tracking Engine
Many institutions use an automated radiology recommendation tracking engine to increase the likelihood of follow- up completion after findings of possible cancer are detected in the abdomen or pelvis. However, these reports often include large free-text segments and are structured differently due to interphysician and interinstitutional variability.
In an article published online in Radiology: Artificial Intelligence, Jackson M. Steinkamp, BA, University of Pennsylvania, Philadelphia, and colleagues developed two neural networks to compare with simple string matching, random forests, extreme gradient boosting and support vector machines to determine if the neural networks can outperform other approaches on organ-level pathology report classification with no domain-specific feature engineering. Results showed that both convolutional neural networks and long shortterm memory networks algorithms achieved approximately 95% accuracy and F1 scores on a highly varied corpus of multi-institution multiformat reports. Researchers also used methods for interpretability from the literature to evaluate the sensibility of the models’ learned parameters and found that the system qualitatively learned features similar to human annotators. “End-to-end neural network architectures perform well on a clinical text-classification task with high levels of human interpretability. Such systems have the potential to improve information extraction and summarization in a wide variety of clinical contexts, toward the ultimate end of improving care quality and efficiency,” the authors write.
Ischemic Infarction in Young Adults: A Review for Radiologists
Ischemic stroke in young adults is devastatingly debilitating and increasingly frequent. The consequences of this familiar disease in this atypical age group are especially detrimental and long lasting. Ischemic stroke in young adults is now emerging as a public health issue, one in which radiologists can play a key role.
In an article published online in RadioGraphics, Jennifer L. McCarty, MD, University of Texas Health Sciences Center at Houston, and colleagues reviewed the epidemiology, workup and causes of ischemic infarction in young adults, focusing on how it differs from that in older adults.
“No universal stroke imaging algorithm exists for young adults, and protocols vary between institutions. Any imaging pathway used for ischemic infarction in young adults should, at the very least, determine the presence or absence of hemorrhage, evaluate for stroke mimics and assess candidacy for intra arterial therapy,” the authors wrote.
Complete Free-breathing Adenosine Stress Cardiac MRI Using Compressed Sensing and Motion Correction: Comparison of Functional Parameters, Perfusion, and Late Enhancement with the Standard Breath-holding Examination
Despite the growing use of cardiac MRI over the past several years, the technique has faced some challenges. Standard cardiac MRI and stress cardiac MRI are demanding examinations with multiple breath-hold (BH) commands and long examination times. Moreover, image quality greatly depends on heart rate and heart rate variability, thereby limiting cardiac MRI, especially in patients with arrhythmia. Thus, reduction of imaging time and application of free-breathing (FB) sequences is desirable to achieve diagnostic image quality in critically ill patients.
In an article published online in Radiology: Cardiothoracic Imaging, Christoph Treutlein, MD, University of Erlangen, Germany, and colleagues compared image quality and diagnostic confidence of functional analysis, stress and rest perfusion and late gadolinium enhancement (LGE) imaging during BH and FB stress cardiac MRI.
Stress cardiac MRI performed with compressed sensing and in-line motion correction delivers diagnostic image quality. Results for functional analysis in FB correlate very well with results in BH. Image quality in BH is superior to that in FB for cine steady-state free precession imaging but is comparable for stress and rest and LGE imaging.
“Stress FB cardiac MRI yields diagnostic image quality and represents an alternative for patients who are unable to meet the demands of multiple BHs and long examination times,” the authors conclude.
Listen to Radiology Editor David A. Bluemke, MD, PhD, discuss this month’s research you need to know. Podcasts summarize the importance and context of selected recent articles. Subscribe today at RSNA.org/Radiology-Podcasts and never miss a single episode.
“A Roadmap for Foundational Research on Artificial Intelligence in Medical Imaging: From the 2018 NIH/RSNA/ACR/The Academy Workshop,” Langlotz C, et al.
“CT of Kidney Volume in Autosomal Dominant Polycystic Kidney Disease: Accuracy, Reproducibility, and Radiation Dose,” Bevilacqua M, et al.
MRI Evaluation of Lymphatic Abnoralities in the Neck and Thorax after Fontan Surgery: Relationship with Outcome,” Biko DM et al.
Listen to RadioGraphics Editor Jeffrey S. Klein, MD, and authors discuss the following articles from recent issues of RadioGraphics at RSNA.org/RG-Podcasts.
"Vascular Anomalies of the Pediatric Liver,” Albers BK, et al.
“Multimodality Imaging of Focal and Diffuse Fibrosing Mediastinitis,” Garrana SH, et al.
Audio summary podcasts (also available on iTunes and Google Play) include these studies:
“Recognizing and Minimizing Artifacts at CT, MRI, US, and Molecular Imaging,” Triche BL, et al.
“Implantable Electronic Stimulation Devices from Head to Sacrum: Imaging Features and Functions,” Sterman J, et al.