A rare presentation of Rosai-Dorfman disease involving the central nervous system.

A 29-year-old gentleman diagnosed with Rosai-Dorfman disease (RDD) on corneal biopsy, 2 years ago, presented with fluctuating left-sided numbness, intermittent slurred speech, and urinary incontinence, progressively worsening over the past three months.

Infectious Diseases of the Brain and Spine: Parasitic and Other Atypical Transmissible Diseases.

Atypical infections of the brain and spine caused by parasites occur in immunocompetent and immunosuppressed hosts, related to exposure and more prevalently in endemic regions. In the United States, the most common parasitic infections that lead to central nervous system manifestations include cysticercosis, echinococcosis, and toxoplasmosis, with toxoplasmosis being the most common opportunistic infection affecting patients with advanced HIV/AIDS. Another rare but devastating transmittable disease is prion disease, which causes rapidly progressive spongiform encephalopathies. Familiarity and understanding of various infectious agents are a crucial aspect of diagnostic neuroradiology, and recognition of unique features can aid timely diagnosis and treatment.

Infectious Diseases of the Brain and Spine: Fungal Diseases.

Advances in treatments of autoimmune diseases, acquired immunodeficiency syndrome, organ transplantation, and the use of long-term devices have increased the rates of atypical infections due to prolonged immune suppression. There is a significant overlap in imaging findings of various fungal infections affecting the central nervous system (CNS), often mimicking those seen in neoplastic and noninfectious inflammatory conditions. Nonetheless, there are imaging characteristics that can aid in distinguishing certain atypical infections. Hence, familiarity with a wide range of infectious agents is an important part of diagnostic neuroradiology. In this article, an in-depth review of fungal diseases of the CNS is provided.

Follow-up infarct volume on fluid attenuated inversion recovery (FLAIR) imaging in distal medium vessel occlusions: the role of cerebral blood volume index.

BACKGROUND: Distal medium vessel occlusions (DMVOs) contribute substantially to the incidence of acute ischemic strokes (AIS) and pose distinct challenges in clinical management and prognosis. Neuroimaging techniques, such as Fluid Attenuation Inversion Recovery (FLAIR) imaging and cerebral blood volume (CBV) index derived from perfusion imaging, have significantly improved our ability to assess the impact of strokes and predict their outcomes. The primary objective of this study was to investigate relationship between follow-up infarct volume (FIV) as assessed by FLAIR imaging in patients with DMVOs. METHODS: This prospectively collected, retrospective reviewed cohort study included patients from two comprehensive stroke centers within the Johns Hopkins Medical Enterprise, spanning August 2018-October 2022. The cohort consisted of adults with AIS attributable to DMVO. Detailed imaging analyses were conducted, encompassing non-contrast CT, CT angiography (CTA), CT perfusion (CTP), and FLAIR imaging. Univariable and multivariable linear regression models were employed to assess the association between different factors and FIV. RESULTS: The study included 79 patients with DMVO stroke with a median age of 69 years (IQR, 62-77 years), and 57% (n = 45) were female. There was a negative correlation between the CBV index and FIV in a univariable linear regression analysis (Beta = - 16; 95% CI, - 23 to - 8.3; p < 0.001) and a multivariable linear regression model (Beta = - 9.1 per 0.1 change; 95% CI, - 15 to - 2.7; p = 0.006). Diabetes was independently associated with larger FIV (Beta = 46; 95% CI, 16 to 75; p = 0.003). Additionally, a higher baseline ASPECTS was associated with lower FIV (Beta = - 30; 95% CI, - 41 to - 20; p < 0.001). CONCLUSION: Our findings underscore the CBV index as an independent association with FIV in DMVOs, which highlights the critical role of collateral circulation in determining stroke outcomes in this patient population. In addition, our study confirms a negative association of ASPECTS with FLAIR FIV and identifies diabetes as independent factor associated with larger FIV. These insights pave the way for further large-scale, prospective studies to corroborate these findings, thereby refining the strategies for stroke prognostication and management.

Sinonasal (Schneiderian) Tumors in the Temporal Bone: Case Series and Systematic Review.

BACKGROUND: Neoplasms derived from the sinonasal epithelium are a rare finding in the temporal bone, and their origins are controversial. PURPOSE: To review the characteristics of sinonasal epithelial (previously known as Schneiderian) tumors occurring in the temporal bone. DATA SOURCE: This was a 2-center case series and systematic review of MEDLINE, EMBASE, and the Web of Science through May 2021. STUDY SELECTION: Patients with clinicopathologic evidence of temporal bone involvement by neoplasms of sinonasal epithelial origin were selected, with or without a history of prior primary sinonasal epithelial tumors. DATA ANALYSIS: Clinical, radiologic, and pathologic data were extracted. DATA SYNTHESIS: The systematic review included 56 studies and our 8 unpublished cases, totaling 76 cases of papillomas or squamous cell carcinomas in the temporal bone. Of these, 51% occurred secondary to sinonasal tumors, and 49% occurred primarily. Secondary tumors were usually metachronous (77%), with a median delay of 1 year from sinonasal-to-temporal bone tumor diagnosis. Most cases were unilateral (90%); bilateral temporal bone involvement occurred only as secondary ("trilateral") tumors. Unilateral secondary tumors had ipsilateral (81%) or bilateral (19%) sinonasal counterparts. Secondary tumors were more likely to be malignant (OR, 6.7, P < .001). LIMITATIONS: The review was based on case reports and small case series, which are subject to reporting bias. CONCLUSIONS: The observed tumor patterns support the hypothesis that the Eustachian tube facilitates the spread of sinonasal epithelium-derived neoplasms from the sinonasal cavity to the temporal bone. Transtubal spread of sinonasal epithelium-derived neoplasms should be considered among the rare causes of middle ear masses.

Not without Context-A Multiple Methods Study on Evaluation and Correction of Automated Brain Tumor Segmentations by Experts.

RATIONALE AND OBJECTIVES: Brain tumor segmentations are integral to the clinical management of patients with glioblastoma, the deadliest primary brain tumor in adults. The manual delineation of tumors is time-consuming and highly provider-dependent. These two problems must be addressed by introducing automated, deep-learning-based segmentation tools. This study aimed to identify criteria experts use to evaluate the quality of automatically generated segmentations and their thought processes as they correct them. MATERIALS AND METHODS: Multiple methods were used to develop a detailed understanding of the complex factors that shape experts' perception of segmentation quality and their thought processes in correcting proposed segmentations. Data from a questionnaire and semistructured interview with neuro-oncologists and neuroradiologists were collected between August and December 2021 and analyzed using a combined deductive and inductive approach. RESULTS: Brain tumors are highly complex and ambiguous segmentation targets. Therefore, physicians rely heavily on the given context related to the patient and clinical context in evaluating the quality and need to correct brain tumor segmentation. Most importantly, the intended clinical application determines the segmentation quality criteria and editing decisions. Physicians' personal beliefs and preferences about the capabilities of AI algorithms and whether questionable areas should not be included are additional criteria influencing the perception of segmentation quality and appearance of an edited segmentation. CONCLUSION: Our findings on experts' perceptions of segmentation quality will allow the design of improved frameworks for expert-centered evaluation of brain tumor segmentation models. In particular, the knowledge presented here can inspire the development of brain tumor-specific metrics for segmentation model training and evaluation.

Simulation Training in the Management of Adverse Contrast Reactions: A Systematic Review.

BACKGROUND: Simulation-based training has become increasingly prominent within medical education, but its utility within radiology remains underexplored. OBJECTIVE: To appraise the evidence for the effectiveness of simulation on the management of adverse reactions to contrast media. METHODS: A systematic search of the literature was conducted. Eligible studies recruited radiology residents, provided simulation-based training focused on contrast reaction management, and measured any effectiveness outcome compared with any nonsimulation training or no training. The quality of studies was appraised and outcomes were classified according to Kirkpatrick's hierarchy and the strength of evidence. RESULTS: Out of 146 screened results, 15 articles were included that described 17 studies-3 randomized trials and 14 pretest-posttest studies of hands-on or, less commonly, computer-based simulation. In all 16 studies that assessed knowledge before and after intervention, written test scores improved after simulation. Most studies noted improvements in comfort or confidence managing contrast reactions as well. In all three studies that assessed knowledge after simulation and after didactic lecture as a control, posttest scores were not statistically significantly better in the simulation groups than the lecture groups. Common study limitations included single-group designs, measuring only learning outcomes using unvalidated instruments, modest sample sizes, and limited assessment of long-term retention. CONCLUSION: Simulation produces subjective improvements and knowledge gain relevant to contrast reaction management. Further research is required to demonstrate superiority of simulation-based contrast reaction management training over traditional didactic lecture-based instruction.

Transformer versus traditional natural language processing: how much data is enough for automated radiology report classification?

OBJECTIVES: Current state-of-the-art natural language processing (NLP) techniques use transformer deep-learning architectures, which depend on large training datasets. We hypothesized that traditional NLP techniques may outperform transformers for smaller radiology report datasets. METHODS: We compared the performance of BioBERT, a deep-learning-based transformer model pre-trained on biomedical text, and three traditional machine-learning models (gradient boosted tree, random forest, and logistic regression) on seven classification tasks given free-text radiology reports. Tasks included detection of appendicitis, diverticulitis, bowel obstruction, and enteritis/colitis on abdomen/pelvis CT reports, ischemic infarct on brain CT/MRI reports, and medial and lateral meniscus tears on knee MRI reports (7,204 total annotated reports). The performance of NLP models on held-out test sets was compared after training using the full training set, and 2.5%, 10%, 25%, 50%, and 75% random subsets of the training data. RESULTS: In all tested classification tasks, BioBERT performed poorly at smaller training sample sizes compared to non-deep-learning NLP models. Specifically, BioBERT required training on approximately 1,000 reports to perform similarly or better than non-deep-learning models. At around 1,250 to 1,500 training samples, the testing performance for all models began to plateau, where additional training data yielded minimal performance gain. CONCLUSIONS: With larger sample sizes, transformer NLP models achieved superior performance in radiology report binary classification tasks. However, with smaller sizes (<1000) and more imbalanced training data, traditional NLP techniques performed better. ADVANCES IN KNOWLEDGE: Our benchmarks can help guide clinical NLP researchers in selecting machine-learning models according to their dataset characteristics.

Screening Mammography Utilization Among Women With Visual or Hearing Impairment: Results From the 2019 NHIS.

To evaluate for differences in breast cancer screening among women with visual or hearing impairment, the 2019 National Health Interview Survey was analyzed for mammography use in the past 2 years among women age 50-74, adjusting for demographic characteristics, health care access, and comorbidities. Visual impairment was independently associated with decreased likelihood of recent mammography (odds ratio [OR], 0.71; 95% CI, 0.59-0.85; p < .001). Hearing impairment was not independently associated with mammography use (OR, 0.91; 95% CI, 0.75-1.11; p = .37).

Clinical applications of arterial spin labeling of the intracranial compartment in vascular anomalies-A case-based review.

Arterial spin labeling (ASL) is a magnetic resonance perfusion technique that allows for quantification of cerebral blood flow (CBF) without the use of contrast or radiation. Several applications of ASL have been described in diagnosis of strokes and stroke mimics, intracranial tumors, and other conditions. Various vascular anomalies exhibit specific CBF patterns that correlate with different signal intensities on ASL. In this case-based review, we demonstrate the utility of ASL in diagnosis and surveillance of vascular anomalies in the intracranial compartment.

Arterial spin labeling clinical applications for brain tumors and tumor treatment complications: A comprehensive case-based review.

Arterial spin labeling (ASL) is a noninvasive neuroimaging technique that allows for quantifying cerebral blood flow without intravenous contrast. Various neurovascular disorders and tumors have cerebral blood flow alterations. Identifying these perfusion changes through ASL can aid in the diagnosis, especially in entities with normal structural imaging. In addition, complications of tumor treatment and tumor progression can also be monitored using ASL. In this case-based review, we demonstrate the clinical applications of ASL in diagnosing and monitoring brain tumors and treatment complications.

Clinical application of brain perfusion imaging in detecting stroke mimics: A review.

Stroke mimics constitute a significant proportion of patients with suspected acute ischemic stroke. These conditions may resemble acute ischemic stroke and demonstrate abnormalities on perfusion imaging sequences. The most common stroke mimics include seizure/epilepsy, migraine with aura, brain tumors, functional disorders, infectious encephalopathies, Wernicke's encephalopathy, and metabolic abnormalities. Brain perfusion imaging techniques, particularly computed tomography perfusion and magnetic resonance perfusion, are being widely used in routine clinical practice for treatment selection in patients presenting with large vessel occlusion. At the same time, the utilization of these imaging modalities enables the opportunity to better diagnose patients with stroke mimics in a time-sensitive setting, leading to appropriate management, decision-making, and resource allocation. In this review, we describe patterns of perfusion abnormalities that could discriminate patients with stroke mimics from those with acute ischemic stroke and provide specific case examples to illustrate these perfusion abnormalities. In addition, we discuss the challenges associated with interpretation of perfusion images in stroke-related pathologies. In general, perfusion imaging can provide additional information in some cases-when used in combination with conventional magnetic resonance imaging and computed tomography-and might help in detecting stroke mimics among patients who present with acute onset focal neurological symptoms.