Radiographics最新文献

Postoperative fluid collections are commonly encountered at imaging after lumbar spine surgery. While small collections usually resolve on their own, larger or more complex ones have a limited differential diagnosis. Varied surgical approaches to the lumbar spine may predispose patients to different types of fluid collections, including seromas, hematomas, pseudomeningoceles, abscesses, bone morphogenetic protein (BMP)-related collections, lymphoceles, and urinomas; thus, radiologists should be aware of the different types of anterior and posterior surgical approaches to lumbar spine interbody fusion. Hematomas and abscesses may occur with any surgical approach. Pseudomeningoceles most frequently result after posterior approach surgery, while urinomas and lymphoceles more likely develop after anterior approach surgery due to the close proximity of the surgical corridor to the nephroureteral system or lymphatics, respectively. Surgical implants or biologics used at surgery can lead to BMP-related collections; awareness of any implanted materials intraoperatively can be helpful at the time of interpretation. While MRI is most frequently used for identifying fluid collections in the postoperative spine, CT can serve as an important adjunct. CT myelography can be used to confirm or exclude the presence of a pseudomeningocele but may not be necessary if MRI shows a confirmatory flow jet, confirming leakage of cerebrospinal fluid into the pseudomeningocele. CT enhanced with intravenous contrast material during the excretory phase also allows confirmation of the presence of a urinoma. When the contents of a postoperative fluid collection remain uncertain, aspiration may be necessary. Accurate diagnosis of the type of postoperative collection is extremely useful in guiding patient management and determining prognosis. ^©RSNA, 2025.

Head US is the mainstay of initial neuroimaging in preterm and term neonates and young infants. Echogenic lesions are a commonly encountered finding on US images, with a wide spectrum of underlying causes. These include normal structures and normal variants such as the choroid plexus and cerebellar vermis, normal transmantle white matter tracts, and benign entities that should not be mistaken for disease, such as hyperechoic caudate nuclei and thalamostriate mineralizing vasculopathy. Many pathologic conditions associated with major morbidity and mortality also may appear echogenic on US images. These most commonly include germinal matrix hemorrhage; white matter injury; hypoxic-ischemic encephalopathy; and hemorrhagic, infectious, and ischemic lesions. These lesions may be differentiated by their location at the caudothalamic groove, in the periventricular or deep white matter, diffusely involving the white matter, involving the deep gray matter, in the peripheral parenchyma, or in the cerebellum. Use of a tailored gestational age- and location-based approach combined with salient clinical details and knowledge of the unique vulnerabilities and responses to perinatal stressors in the preterm and term period improve diagnostic confidence and enable radiologists to differentiate those echogenic lesions that require no further follow-up from those that require further imaging or laboratory workup and specialist referral. ^©RSNA, 2025 Supplemental material and the slide presentation from the RSNA Annual Meeting are available for this article.

Transthoracic needle biopsy (TTNB) is a well-established method of diagnostic evaluation used to answer a range of clinical questions and is an important diagnostic tool in the workup and management of patients with suspected or known malignancy or infection. Over the past decade, the body of literature on TTNB has continued to elucidate and support best practices and shed light on emerging uses of TTNB such as molecular testing. As such, TTNB should be considered when technically feasible and safe. Developing and maintaining an evidence-based approach to preprocedural evaluation, procedural technique, risks or risk mitigation, and practice-related controversies helps the radiologist optimize procedural outcomes and serve as an effective consultant to multidisciplinary teams that seek out this procedure to help guide patient care. ^©RSNA, 2025.

Pancreas transplant (PT) improves quality of life and longevity in patients with diabetes by restoring endocrine function, eliminating insulin use, and reducing long-term complications. The benefits of PT must be weighed against the risks, including postsurgical complications and long-term immunosuppression. Surgical advances and improved immunosuppression regimens in the past 2 decades have led to improved patient and allograft survival rates. Imaging a PT can be challenging because of the complex postsurgical anatomy and difficulty distinguishing the PT from adjacent bowel, particularly at US. Nevertheless, US is the first-line imaging modality for evaluating the allograft and vasculature for complications. Contrast-enhanced US is a useful next step to evaluate blood flow to the allograft when Doppler US findings are inconclusive. CT with oral contrast material is preferred for evaluating suspected bowel complications and is useful for evaluating fluid collections. CT angiography is a rapid way to evaluate vasculature and bleeding complications if iodinated contrast material is not contraindicated. MRI can be used as a problem-solving tool to further evaluate complications, particularly when iodinated contrast material is contraindicated or radiation risk is a concern. Ferumoxytol-enhanced MRI is advantageous for evaluating transplant vasculature because of the prolonged intravascular state of the contrast material. It is important for radiologists to understand the spectrum of normal imaging appearances of the pancreas allograft, as well as various posttransplant complications, to provide optimal care. The authors discuss surgical techniques and the anatomy of PT, review imaging evaluation, and describe complications. ^©RSNA, 2025 Supplemental material is available for this article.

Targeted radionuclide therapy (TRT) plays an important role in management of oncology patients, particularly those with thyroid cancer, prostate cancer, or neuroendocrine tumor. Use of TRTs has expanded rapidly in recent years, with approval of new agents and indications. The number of ongoing TRT clinical trials suggests that this trend is likely to continue in the future. Adequate knowledge of potential adverse events and their management is essential for treating physicians to optimize outcomes in patients undergoing TRT. The authors review adverse events that may occur in TRT; their clinical manifestations, incidence, risk factors, and management; and strategies to prevent or decrease their risk of occurrence. ^©RSNA, 2025.

In radiology practice, medical images are described and interpreted by radiologists in text reports. Recent technical developments enabling deep learning models to connect images and text may facilitate the radiologic workflow. These developments include advances in data embedding, self-supervised learning, zero-shot learning, and transformer-based model architectures. Models connecting images and text can be divided into four categories: (a) Text-image alignment models associate text descriptions with corresponding images. (b) Image-to-text models create text descriptions from images. (c) Text-to-image models generate images from text descriptions. (d) Multimodal models integrate and interpret multiple types of data such as images, videos, text, and numbers simultaneously. Potential clinical applications of these models include automated captioning of medical images, generation of the preliminary radiology report, and creation of educational images. These advances may enable case prioritization, streamlining of clinical workflows, and improvements in diagnostic accuracy. Published under a CC BY 4.0 license.

Health disparities represent a significant and growing concern in the United States, largely due to recent recognition of worse outcomes experienced by historically underserved groups. Radiologists have an opportunity to address these disparities by using community engagement (CE) to deepen their knowledge about patient and community-level barriers and inform the development of programs focused on improving patient access to timely radiology care. Various established frameworks, such as the socioecologic model and the National Institute on Minority Health and Health Disparities research framework, can be adapted and leveraged by radiology practices to promote CE and advance health equity. These models emphasize multilevel interactions among individual factors to broader societal influences, enabling tailored approaches to addressing disparities in access to radiologic services. Providing practical ways of applying these models will increase the presence and impact of radiology in the community. In this article, case-based practical examples of current CE efforts in radiology are presented, with the aim of empowering radiologists to engage in and lead community-focused initiatives. For each case study, specific disparities and access to radiologic services, as well as intervention strategies, are described. Finally, the authors present adaptations to existing models and frameworks to inform a radiology-centric CE model that can be used to sustainably address health disparities by enhancing educational efforts, building coalitions, and developing comprehensive outreach strategies. This proposed model is intended to not only alleviate disparities but also mitigate radiologist burnout, thereby enhancing both community health outcomes and radiologist well-being. ^©RSNA, 2025 See the invited commentary by Hajjar and Brewington in this issue.