The practice of paying referral fees, in which radiology service providers offer a percentage of the cost of radiology tests to referring clinicians, is known to exist in India; however, the extent of this practice is not well documented. This study aims to explore how common is this practice and how it affects the cost and quality of radiology services. The study also investigates if this practice varies between the private hospitals and diagnostic centers. An anonymous online survey was distributed between November 2024 and January 2025 to radiologists currently practicing in India. Total 440 responses were received and the survey data were analyzed using descriptive and comparative statistics. 85.88% radiologists have witnessed referral fee payment and 80.6% of radiologists believe that at least 20% of cost of any radiology investigation is paid as commission. 58.55% radiologists are of view that referral fee incentivizes clinicians to order unnecessary radiology investigations. Majority of radiologists think that kickback both increases the cost of radiology services for patients and reduces the net radiology revenue per test, which in turn compels them to perform more investigations in shorter timeframe. 75.28% radiologists believe that referral fee reduces the quality of radiology services. Referral fee is more common in diagnostic centres in comparison to the hospitals. As per perception of practicing radiologists, referral fee payment for diagnostic radiology tests is very common practice in India, which prompts clinicians to order unnecessary radiology tests, increases costs for patients, adds to burnout of radiologists, ultimately negatively impacting the quality of radiology services.
Musculoskeletal (MSK) disorders are among the leading contributors to disability in the United States, and are inherently diagnosis-dependent, relying on radiologic evaluation to define pathology, stratify disease severity, and guide escalation of care. Although substantial interstate variation in MSK disability burden has been reported, it is unclear whether these geographic differences represent transient trends or persistent structural patterns that may reflect inequities in diagnostic access and care delivery. Understanding whether high-burden states are improving, stagnating, or falling further behind is critical for radiology-relevant health policy, imaging workforce planning, and equitable resource allocation. This study used the Global Burden of Disease (GBD) 2010-2023 age-standardized Disability-Adjusted Life Year (DALY) rates for all MSK causes combined. Cross-sectional MSK burden was defined using a 3-year average (2021-2023). States were grouped into tertiles-low, intermediate, and high-burden-based on this average. Long-term temporal changes were assessed using the Compound Annual Growth Rate (CAGR) from 2010 to 2023. Temporal stability was evaluated through a heatmap of annual DALY values. High-burden states had an average of 4,230 MSK DALYs per 100,000, compared with 3,609 per 100,000 in low-burden states, representing a 620 DALY (17 %) difference. Although most states showed modest declines in MSK burden, improvement was minimal overall (mean compound annual growth rate (CAGR) -0.16 % per year). High-burden states showed the least improvement in DALYs (-0.05 % per year), compared with faster declines in low-burden states (-0.26 % per year), and seven high-burden states experienced a worsening burden over time. Heatmap visualization demonstrated highly persistent geographic patterns, with states rarely shifting between burden categories over the 14-year period. States with the highest MSK disability burden are improving the slowest, resulting in a widening MSK disability gap. Because MSK care is dependent on imaging for diagnosis and treatment escalation, these disparities may suggest potential disparities in access to timely MSK imaging and subspecialty expertise. Radiology can directly contribute to reducing long-term MSK disability by aligning imaging capacity and workforce resources with regions experiencing the highest burden.
Disparities in research grant funding have been documented across various fields of medicine; yet gender inequities within radiology remain underexplored. To evaluate gender-based differences in NIH radiology grant funding, focusing on the number of grants awarded and average funding amounts received by principal investigators (PIs) from 2009 to 2023. A retrospective cross-sectional analysis was conducted using publicly available NIH RePORTER data. Grants awarded to PIs affiliated with radiology departments from 2009 to 2023 were included. Gender was inferred using Genderize.io and verified manually. Extracted data included number of grants, average and total funding amounts, and PI gender. Analyses included descriptive statistics, t-tests for comparing means, and linear regression to evaluate temporal trends. A total of 9,378 NIH-funded radiology grants were analyzed. Of these, 75.2% (n=7,056) went to men and 24.8% (n=2,322) to women. Men received significantly higher average funding per grant than women ($704.2 K vs $535.5 K; p < 0.0001) and were more likely to be repeat recipients (p < 0.05), revealing a persistent funding gap between genders. Total funding to men PIs was $5.0 B versus $1.3 B for women PIs. While the men-to-women ratio narrowed (7:1 in 2009 to 2.8:1 in 2023), men consistently received higher grant volume and funding annually. Despite incremental progress in women's representation, these findings underscore ongoing gender disparities in radiology funding. This imbalance may hinder innovation and limit the range of perspectives driving future research. Evidence-informed strategies may help address inequities, promote diversity and innovation, and ensure equitable, high-quality patient care.
Artificial intelligence (AI) is increasingly integrated into radiology across multiple workflow levels, with its role as a simultaneous second reader holding particular promise. We performed an umbrella review of systematic reviews and meta-analyses reporting pooled diagnostic accuracy of AI models using clinician (including radiologist) interpretation as the reference standard. References were identified through queries of Pubmed, Scopus, Embase, and Google Scholar (last updated January 7th, 2025). Data were analyzed using metaumbrella tool within R statistical software with stratification of evidence by Ioannidis criteria. Study quality was assessed using the AMSTAR-2 tool. From 1,719 unique references, ten meta-analyses met inclusion criteria, encompassing 147 primary studies with over 722,000 case and 3.6 million control images. Diagnostic odds ratios ranged from 30.67 (95% CI; 10.06-102.87), fracture detection on X-ray, to 273.60 (95% CI; 130.51-573.58), pulmonary nodule detection on CT. Most meta-analyses (n = 9) provided Class II evidence, reflecting highly suggestive findings limited by invariably substantial heterogeneity (I² = 89.9%-99.9%). The quality was assessed as critically low in nine reviews and low in one. AI models have shown strong diagnostic performance across various radiologic applications. Due to our inclusion criteria requiring clinician/radiologist interpretation as the reference standard, these findings reflect AI-human agreement rather than AI accuracy using a more definitive ground truth (e.g. histopathology). Furthermore, the strength of this evidence is limited by substantial heterogeneity, variability in imaging modalities, and differences in model development and validation.
Diversity within medical training enhances academic development and patient care. The residency recruitment and selection process plays a critical role in shaping workforce diversity, yet no comprehensive framework exists to systematically evaluate these processes. This study applies a novel conceptual framework to assess the gender distribution of applicants to our institution's Diagnostic Radiology residency program before and after the transition from in-person to virtual interviews. We analyzed ERAS data from 2017 to 2023, examining gender distribution at each stage of the recruitment and selection process: application, interview selection, ranking, and matching. Statistical analyses included chi-square tests and cumulative logistic regression to compare outcomes between in-person (2017-2020) and virtual (2021-2023) interview cycles. The proportion of female applicants, interviewees, and ranked candidates remained stable between in-person and virtual interviews. However, significantly fewer women matched into the program during virtual interviews (5.6% vs. 37.5%, p = 0.02). There were no significant changes in gender distribution across rank list quartiles. Our novel framework provides a systematic approach to evaluating residency recruitment and selection processes, allowing for a more nuanced interpretation of where disparities emerge. Applying this framework revealed that while selection practices remained consistent, the transition to virtual interviews may have hindered successful recruitment of female applicants. Future studies can leverage this framework to examine additional diversity metrics and guide institutional efforts toward more equitable residency selection.
Predictive machine learning (ML) models may help reduce radiology appointment no-shows and late cancellations, which disrupt care, reduce operational efficiency, and cause financial losses. However, few studies address the practical challenges of developing robust no-show models across large health systems. This study outlines methodological hurdles we faced in building outpatient radiology no-show prediction models at a large academic medical center. The dataset includes 334,002 outpatient imaging appointments across multiple imaging sites. EHR data was integrated with external datasets for weather, socioeconomic indicators, and geographic data. Preprocessing included cohort selection, feature engineering, and curation strategies to address class imbalance and leakage. Boosted tree classifiers and other standard algorithms were trained on 2022-2023 data and tested on 2024 data. Our best models achieved AUC = 0.71, consistent with the low end of prior results. Challenges included ambiguous data definitions, distinguishing cancellations from reschedules, and variable data quality across features and sites. Feature engineering was constrained by predictor granularity, and temporal inconsistencies highlighted the risk of data leakage. Key predictors were phone reminder status, appointment confirmation, and modality, while crime, income, and weather showed little utility. Building no-show models requires extensive data cleaning, feature engineering, and temporal validation. Real-world data constraints increase risks of flawed models from leakage, inconsistencies, and confounding, particularly for prospective predictions. Successful deployment of no-show models requires more than algorithmic sophistication. Transparent documentation, rigorous temporal validation, and critical data quality assessments are essential to ensure models are reliable, fair, and deployable.
To characterize how non-interventional radiology (IR) medical providers discuss IR procedures on X.com (formerly Twitter, Inc; San Francisco, CA), including awareness, referral intent, engagement, and misinformation. This retrospective infodemiology study evaluated public English-language X posts from January 1, 2024, through December 31, 2025. Long-form queries were used to identify posts related to uterine fibroid embolization (UFE), prostatic artery embolization (PAE), Y90 radioembolization (Y90), transarterial chemoembolization (TACE), gastric artery embolization (GAE), tumor ablation, and related procedures. After de-duplication and exclusion of reposts without commentary, IR-authored posts, and likely marketing or spam accounts, the final primary corpus comprised approximately 5,220 posts. Posts were coded for sentiment, awareness level, and thematic content. Statistical analyses included chi-square tests for categorical comparisons, Kruskal-Wallis tests for nonparametric continuous comparisons, Mann-Kendall trend tests for directional volume analysis, and Wilson's method for confidence interval estimation of referral-intent prevalence. Discussion volume increased over the study period, although quarterly trend testing did not reach conventional statistical significance (tau = 0.43; P = .07). PAE and Y90/TACE produced the highest adjusted post volumes (approximately 4,387 and 4,458 posts, respectively), whereas UFE yielded a smaller but more procedure-specific corpus (approximately 466 posts). GAE demonstrated a high raw post volume (n = 13,853) but an estimated on-topic rate of only 12%, reflecting heavy contamination from non-medical acronyms; the adjusted on-topic estimate was approximately 1,662 posts. Urology and radiation oncology providers showed the clearest engagement with PAE-related content, including explicit referral-intent posts. Medical and surgical oncology accounts were active in Y90/TACE discussions, including research and access-barrier themes. No verified obstetrics and gynecology physician accounts were identified in the long-form UFE corpus. Referral-intent posts represented 8.1% of the primary corpus (95% confidence interval, 6.5%-9.8%; approximately 418 posts) and generated greater engagement than routine educational posts. Misinformation represented approximately 7% of posts by volume but achieved disproportionate reach. Awareness of IR procedures on social media was uneven and specialty-dependent, with misinformation achieving disproportionate reach. These findings support targeted outreach to referring specialties and proactive monitoring of promoted information on social media.
Recent advancements in Artificial Intelligence (AI)-driven algorithms have improved patient alignment in Computed Tomography (CT) imaging. However, studies mainly focus on single scanners or specific body areas, indicating a need for broader evaluations. Our study uses a Dose Monitoring System (DMS) to compare vertical shifts in CT exams from two scanners, one AI-based and one manually operated. We analysed 6983 CT scans from 3000 patients on two scanners operated by the same radiology team using the GE DoseWatch (GE Healthcare, Milwaukee, USA) platform. Statistical analysis included tests for normality and distribution comparison (p<0.05). Parameter estimation used an iterative bootstrap method. We also evaluated how many scans have vertical displacement greater than 20 mm. Our results showed non-Gaussian vertical shift distributions for both scanners (p <0.01) and significant differences between them (p < 0.01). Notably, 23% of Ascend exams had vertical shifts beyond ±20 mm, compared to 43% for Lightspeed, indicating substantial improvement with AI-assisted positioning. These findings demonstrate that DMSs can measure alignment accuracy, allowing for the comparison of positioning protocols. However, 23% of AI-assisted examinations showed misalignment, highlighting the need for ongoing training and oversight for technical staff, especially in complex cases. Limitations include reliance on specific software and the absence of image quality and radiation dose comparisons. Future studies should analyse the longterm performance of AI in various clinical settings. This study highlights the importance of continuous data analysis for monitoring system performance and identifying training needs. Better positioning accuracy enhances patient care. The study suggests that while AI-based positioning systems provide substantial benefits, their practical use depends on careful integration into clinical workflows and ongoing training of technical staff.
Patient experience is a core dimension of healthcare quality and is associated with clinical outcomes, safety, and healthcare utilization. At a large academic medical center, nuclear medicine patients reported lower satisfaction with waiting area comfort compared with other radiology patients. Press Ganey patient experience survey data demonstrated a 10.2-percentage-point gap in top-box scores for "comfort in the waiting area" for nuclear medicine patients (76%) compared with other radiology services (86.2%). Using the Six Sigma DMAIC framework, a multidisciplinary team conducted Gemba walks, stakeholder interviews, and cause-and-effect analysis to identify drivers of dissatisfaction. Interventions targeted three key domains: privacy, communication about wait times, and physical comfort. Interventions included relocating cardiac/stress test patients to a private waiting area, implementing pre-visit patient portal messaging to set expectations for wait times, and offering blankets at check-in to address temperature discomfort. The top-box score for waiting area comfort improved from 76% to 85%, representing a 9.0-percentage-point absolute improvement (11.8% relative increase). The balancing measure, "wait time during check-in," remained stable (89% to 90%). Low-cost, workflow-embedded interventions addressing environmental comfort, privacy, and communication significantly improved patient experience without adversely affecting operational efficiency. These findings demonstrate that targeted environmental and process changes can meaningfully enhance patient-centered care in diagnostic imaging settings.
Injuries of the foot and ankle are common in any emergency department. Plain radiographs remain the standard initial imaging offered, with fractures continuing to be the most common type of missed injuries, owing to various pitfalls and limitations, leading to increased use of Computed Tomography (CT) and Magnetic Resonance Imaging (MRI). Although claims related to missed fractures are relatively low volume and of low financial value, they have a financial cost and can have serious and prolonged impact on patients and the staff. This article will focus on the most clinically and medicolegally relevant fractures of the hindfoot, that are at a higher risk of being overlooked, which would be beneficial to imaging interpreters of any grade. Main areas covered include various processes of the talus and calcaneum with a short review of avulsion fractures related to the distal tibiofibular joint. The article will discuss the anatomy, mechanism of injury, provide examples and briefly review the implications of missing such injuries. With litigations related to missed fractures on the rise, it is critical to understand the limitations of plain film, role of early CT/MRI and interpret the plain films with increased awareness of high-risk areas with the maximum incidence of missed injuries. This article would be beneficial for radiologists, inexperienced and experienced clinicians of all grades in raising awareness with a view towards minimising and mitigating the risk of missed and delayed diagnosis of the more commonly overlooked fractures of the hindfoot.
Internal mammary lymph nodes (IMLNs) play a pivotal role in breast cancer staging, prognosis, and therapeutic planning, yet challenges in detection and ongoing controversies regarding their management persist. This review consolidates current evidence on the anatomy, lymphatic function, imaging features, biopsy considerations, and clinical implications of IMLNs. These nodes, visualized in 1-48% of patients depending on imaging modality, serve as secondary lymphatic drainage sites, particularly for medially or deeply located tumors. Metastatic involvement, observed in 1-5% of cases, is classified as N3b according to the AJCC 8th edition. Advanced imaging tools such as ultrasound and MRI facilitate differentiation between benign IMLNs (oval with a central fatty hilum) and malignant ones (round with cortical thickening >3 mm). While ultrasound may assist in detecting suspicious IMLNs, percutaneous biopsy is generally not recommended due to anatomical proximity to major vessels and the risk of complications. Internal mammary sentinel lymph node biopsy (IM-SLNB) offers a minimally invasive staging option, with reported visualization rates of 63-72%. Select studies suggest that IMLN irradiation may reduce recurrence and mortality, although survival benefits remain debated. Emerging predictive models-including nomograms and AI-based radiomics-show promise in preoperative risk stratification. This review highlights the need for standardized diagnostic protocols, routine evaluation of IMLNs in high-risk patients, and large-scale prospective trials to validate the utility of IM-SLNB and optimize therapeutic strategies. Integrating advanced imaging with precision medicine may ultimately enhance detection, tailor treatment, and improve outcomes in breast cancer management.
Supply-related device charges represent a substantial component of interventional radiology (IR) procedure expenditures, yet formal education in charge awareness during training is variable. This study assessed baseline knowledge of supply-related procedure and device charges among IR residents and attendings and evaluated a targeted educational intervention for senior residents. IR residents and attendings at a single academic medical center completed a survey assessing attitudes toward cost-conscious care and estimating supply-related charges for 10 common IR procedures and 8 devices. Senior residents (PGY-5/6) received an intervention consisting of real-time feedback of total supply charges after procedures, instruction on accessing itemized charge data within the electronic medical record, and two didactic sessions addressing device economics and institutional charge variation. The survey was repeated after four months. Accuracy was defined as estimates within ±25% of reference charges. Pre- and post-intervention responses were compared by training level. Response rates were 23/27 (85%) pre-intervention and 20/27 (74%) post-intervention. At baseline, all respondents endorsed the importance of cost-conscious practice and a desire to learn device charges; only 1/23 (4%) reported prior dedicated training. Among senior residents, those reporting adequate education about device charges increased from 12% to 86% (P = 0.010), and perceived access to institutional charge information increased from 0% to 71% (P = 0.007). Procedure charge accuracy improved from 14% to 33% (P = 0.006), and device charge accuracy improved from 4.7% to 29% (P < 0.001). IR trainees and attendings demonstrated limited baseline knowledge of supply-related charges. A low-resource educational intervention improved senior resident awareness and estimation accuracy.
To conduct an update of a prior survey and contribute contemporary social media usage statistics to the radiology education literature. We administered an updated 14-question survey on SurveyMonkey and promoted it through our social media accounts and newsletter subscriber lists between March 10 and April 16, 2025. The survey collected demographic information and was open to the broader medical community, asking respondents to specify their role in medicine and their use of each social media platform for radiology education. We received 320 responses. Of the 318 respondents who specified their role in medicine, 160 were physicians, 65 were technologists, 63 were residents or fellows, 9 were medical students, 3 were nurses, 1 was an administrator, and 17 identified as "other". A strong majority (n = 286, 89.9%) reported working primarily in radiology. Respondents represented 73 countries, most commonly the United States (n = 69, 21.6%), India (n = 34, 10.6%), and Romania (n = 11, 3.4%). The most common age groups were 25-34 years (30.9%), 35-44 years (20.9%), and 45-54 years (17.8%). Most respondents (80.0%) reported using social media for radiology education "every day" or "a few times a week." YouTube was the most frequently used platform (73.1%), followed by Instagram (49.4%) and Facebook (39.1%). YouTube was also the most commonly preferred platform (37.2%). The most frequently reported goals for using social media included "to learn more in radiology" (84.4%) and "to keep updated on current events in radiology" (69.1%). Overall, 74.4% rated social media as "extremely" or "very" useful for radiology education, and 46.6% reported learning more from social media than from traditional resources. In this updated global survey, social media remains widely and frequently used for radiology education, and most respondents consider it highly useful. Patterns of platform use have shifted, with emerging platforms beginning to play more prominent roles while some traditional platforms show declining use. These evolving trends highlight the importance of understanding which platforms different groups use so that educators can target content appropriately and underscore the need for periodic reassessment of social media in radiology education.
Differences in severity and complexity among polytrauma patients remain a major challenge for clinicians and radiologists worldwide. We aimed to provide a descriptive analysis of patient characteristics, trauma mechanisms, and severity in connection with the need for intensive care in polytrauma patients undergoing computed tomography (CT). This retrospective monocentric analysis consecutively included 1993 patients (1305 males, mean age 50.67±23.40 years) with 2002 emergency room cases between 12/2015 and 06/2021. Nine patients were examined twice during the study period. CT was performed in 1897/2002 cases (94.8%). Additional magnetic resonance imaging (MRI) was obtained in 438/2002 cases. For subgroup analysis, trauma mechanisms were compared. Besides conventional trauma causes also critically ill non-trauma patients (CINT) were included. Admission time, day of the week, and month were also analyzed. Traffic accidents constituted the largest trauma group, accounting for 847/2002 cases (42.3%). CINT with non-traumatic emergency died significantly more often (p < 0.001) than patients within the other trauma subgroups. In patients with an injury at the neck area, significantly more often an intubation was registered (p < 0.001). MRI was significantly more frequently performed in cases involving injuries to the cranium (p < 0.001), face (p < 0.001), or neck (p < 0.001). Mean time spent in the hospital was 9.96±13.49 days, and the mean time in the intensive care unit (ICU) was 6.58±9.67 days. Overall, 1171/2002 patients (58.5%) required at least one night in the ICU. A total of 233 patients died following trauma (11.6%), with a mean age of 69.92±18.67years. Imaging strategies in polytrauma patients vary depending on the mechanism of injury and the affected body region. Especially additional MRI was more often needed when brain, face, or neck trauma was present.
To compare prostate-specific membrane antigen (PSMA) PET/CT and multiparametric MRI (mpMRI) for detecting locally recurrent prostate cancer following high-intensity focused ultrasound (HIFU). This retrospective, single-center study included 31 patients treated with HIFU who subsequently underwent both mpMRI and ^18F-piflufolastat PSMA PET/CT followed by prostate biopsy. Imaging studies were independently interpreted by a blinded abdominal radiologist. Diagnostic performance metrics were calculated for each modality and for combined imaging, using biopsy as the reference standard. Paired comparisons were performed using McNemar tests. Spearman rank correlation was used to evaluate associations between SUVmax, post-treatment PSA levels, and ISUP grade group. Of 31 patients, 25 (80.7%) had biopsy-confirmed recurrent prostate cancer. Sensitivity was 72.0% for PSMA PET/CT and 76.0% for mpMRI, with identical specificity of 83.3% for both modalities. Overall accuracy was 74.2% for PSMA PET/CT and 77.4% for mpMRI. Combined imaging increased sensitivity to 92.0% and accuracy to 87.1%, with reduced specificity (66.7%). Combined imaging demonstrated significantly higher sensitivity than either modality alone (p < 0.05). SUVmax showed a modest positive correlation with ISUP grade group (ρ = 0.39, p = 0.03), while post-treatment PSA levels showed no significant correlation with imaging or histopathology. PSMA PET/CT and mpMRI demonstrate comparable standalone performance for detecting locally recurrent prostate cancer after HIFU. Combined interpretation improves sensitivity and overall accuracy, supporting a complementary role for integrated imaging in post-HIFU surveillance.
With the reinstatement of the American Board of Radiology (ABR) oral board examination, optimal preparation strategies for the reinaugural classes remain undefined. Many radiology faculty, who are the cornerstone of residents' education, lack time availability and prior exposure to the oral exam format, creating a need for efficient, scalable training tools. Case readouts are a daily practice in which radiology residents and faculty collaboratively review unreported imaging studies, serving as the primary means of faculty-to-resident education outside of formal lectures. Large language models (LLMs), such as ChatGPT and OpenEvidence, can simulate human conversation and are revolutionizing education. We demonstrate the utility of LLMs as a framework for simulating oral board cases during readouts. Models can be created incorporating ABR preparation materials and the established one-minute preceptor technique to enable structured, efficient use in clinical workflows. We present custom instructions that can be copied into ChatGPT, OpenEvidence, or the preceptor's preferred LLM and are readily accessible for faculty to utilize in daily routine resident readouts. The aspired objective is for the "Radiology Oral Board Coach" to assist the busy radiology faculty in preparing future oral board candidates and ideally enhance residents' education and dictation proficiency.
This study aimed to evaluate the clinical presentation and imaging characteristics of atypical retroperitoneal paragangliomas using contrast-enhanced computed tomography (CECT), and ultrasonography (US). Additionally, the study sought to identify imaging features that facilitate early and accurate preoperative diagnosis. A total of 69 patients (46 male; mean age, 52.0 years) with pathologically confirmed retroperitoneal paragangliomas, diagnosed between May 2015 and February 2025, were included. All patients underwent CECT, and 45 underwent US. Images were independently reviewed by two experienced radiologists, with clinical presentations and diagnostic characteristics systematically recorded. Based on the intraoperative hemodynamic status, patients were divided into two groups: the intraoperative hemodynamic instability group and the intraoperative hemodynamic stability group. This study analyzed the relationships between clinical characteristics as well as imaging features and the intraoperative hemodynamic status. CECT demonstrated that most retroperitoneal paragangliomas appeared as oval-shaped or lobulated masses. The majority of these tumors exhibited rapid washout and minimal washout patterns, along with avid enhancement and the presence of necrosis. They were typically located adjacent to the abdominal aorta and inferior vena cava. US revealed that these tumors presented as hypoechoic, oval-shaped or lobulated, heterogeneous masses, with central anechoic areas being commonly observed, and they were also situated in proximity to the abdominal aorta and inferior vena cava. RPPGs exhibiting a minimal washout or persistent enhancement pattern on CECT exhibited a significantly higher risk of intraoperative hemodynamic instability than those with rapid washout pattern (P = 0.021). There was no statistically significant difference between the clinical features and the remaining CECT imaging features between patients with intraoperative hemodynamic instability and stability. Distinct diagnostic features identified through CECT and US may contribute to the accurate diagnosis of atypical retroperitoneal paragangliomas.
Early hands-on training in image-guided procedures is limited during radiology training. We described our institutional experience implementing cadaveric simulation for CT- and ultrasound (US)-guided procedures to address limited procedural exposure and trainee anxiety early in fellowship. A single-day cadaveric simulation program was implemented for incoming abdominal and musculoskeletal imaging fellows, providing supervised hands-on practice in multiple CT- and US-guided procedures using clinical equipment. Cadaveric simulation was feasible and associated with significant reductions in trainee-reported anxiety and significant improvements in procedural familiarity and confidence across all assessed domains (all p < 0.05). Cadaveric simulation may serve as a valuable adjunct to early procedural training, with future work needed to evaluate objective clinical performance outcomes.
To evaluate the impact of expanding radiologic technologist aide responsibilities on procedural access in the interventional radiology (IR) department of a large academic medical center during radiologic technologist shortages. This quality improvement initiative was conducted in IR between October 1, 2024, and September 30, 2025, following closure of a fluoroscopic procedure room due to radiologic technologist shortages. Radiologic technologist aides were trained in sterile technique and ultrasound setup to assist with bedside and non-fluoroscopic procedures performed under Advanced Practice Provider (APP) or physician supervision. Procedural data were collected retrospectively and filtered to isolate cases performed in one of our IR suites or bedside. Outcomes included case volume, radiologic technologist aide involvement, and safety. Of 16,841 IR procedures, 1,048 were included, and radiologic technologist aides participated in 205 cases (19.6 %). Radiologic technologist aide involvement was most frequent in ultrasound-guided procedures (147/205, 71.7 %), followed by non-image-guided (37/205, 18.0 %) and fluoroscopic cases (21/205, 10.2 %), the latter always with technologist pairing. Among APP-led cases, radiologic technologist aides most commonly supported ultrasound-guided drainage procedures (69/115, 60 %) and non-image-guided tunneled catheter or port removals (34/115, 29.6 %). Among physician-led cases, radiologic technologist aides most frequently supported ultrasound-guided venous ablation/sclerotherapy (53/90, 58.9 %) and drainage procedures (20/90, 22.2 %). No radiologic technologist aide attributable complications occurred. Expanding radiologic technologist aide responsibilities supported IR services in a procedure room and bedside setting during radiologic technologist shortages. Training radiologic technologist aides to assist with sterile preparation and ultrasound-guided workflows helped preserve procedural access to care.
Academic radiology departments in the United States are essential to patient care, education, research, and innovation, yet they face growing challenges such as staffing shortages, rising clinical demands, limited protected academic time, and difficulties retaining residents and junior faculty. This report discusses the current challenges in academic radiology and explores strategies for success and sustainability through the perspectives of early- to mid-career yet accomplished faculty members across core academic domains: education and administration, research, innovation, and artificial intelligence. Drawing from in-depth interviews with leaders at a large tertiary center, we highlight their personal journeys, approaches and initiatives that have fostered career development, interdisciplinary collaboration, and a culture of mentorship. Key themes include the importance of passion-driven leadership, flexible career paths, clinical excellence as a foundation for impactful research, and the need to create environments that nurture innovation and empower emerging talent. By sharing actionable insights and real-world experiences, this report offers a framework for how academic radiologists can navigate current challenges and build a more sustainable future-one that supports academic productivity, professional fulfillment, and the next generation of radiology leaders.