Canadian Association of Radiologists Journal-Journal De L Association Canadienne Des Radiologistes最新文献

Background: Call is among the most challenging aspects of radiology residency, fostering both knowledge development and decision-making under pressure. As imaging volumes have increased, call conditions have evolved. However, little is currently known about the variability in call structures among institutions. This study explores the perceptions of resident call across Canadian radiology programs.

Methods: A 28-question, bilingual, anonymous survey was distributed to all Canadian radiology residents via the Canadian Association of Radiologists (CAR) and institutional representatives of the CAR Resident and Fellow Section (RFS). Anonymous data was collected including demographic information, sleep habits, staff supervision, interactions with referring physicians, and wellness.

Results: Our survey was open from Sept 1, 2024, to Feb 28, 2025. Altogether, 112 residents responded from all training levels and programs, with 71.4% completing it in English and 28.6% in French. Most PGY-2 residents reported 8 to 11 call shifts in the preceding 8 weeks, compared to 4 to 7 shifts among PGY-3 to PGY-5. Close to half of residents reported that they did not sleep at all on call. Among those who did, 70.3% of them reported moderately insufficient sleep. Interactions with referring physicians were poor, with 49.2% of residents rating their experiences as unsatisfactory or very unsatisfactory. Lower call satisfaction was significantly associated with greater call volume and reduced sleep.

Conclusions: Canadian radiology residents reported high call volumes, limited rest, and challenging interactions with referring physicians, all of which were linked to lower satisfaction and well-being. These findings highlight opportunities to optimize call structures, strengthen supervision, and improve resident wellness.

Purpose: Breast arterial calcification (BAC), detectable on routine mammograms, offers a promising independent risk factor for cardiovascular disease (CVD) risk stratification. However, current BAC assessment methods lack standardization and rely on subjective interpretations. This study introduces a semi-supervised deep learning (DL) model to automate BAC severity grading, enhance cross-system generalizability, and align with clinical consensus.

Methods: A U-Net-based segmentation model was trained on 2560 annotated screening mammograms from 7 vendors. A semi-supervised learning strategy employing progressive pseudo-labeling incorporated 6000 unlabeled images to enhance model robustness. BAC severity was graded by thresholding the percentage area covered by BAC and benchmarked against radiologists' assessments using Canadian Society of Breast Imaging (CSBI) guidelines. Performance was evaluated using the Jaccard Similarity Coefficient (JSC) for segmentation, along with accuracy, precision, F1-score, and recall. For detecting clinically significant (Grade 3) BAC, sensitivity, specificity, and area under the curve (AUC) were assessed. Agreement with experts was evaluated using weighted kappa statistics.

Results: The proposed model achieved a JSC of 0.614, an accuracy of 0.991, an F1-score of 0.756, a precision of 0.763, and a recall of 0.764. It demonstrated superior segmentation accuracy compared to the baseline U-Net model. Agreement with consensus radiologists was high, with a weighted kappa of 0.90, 95% CI = (0.70, 1.00). For clinically significant (Grade 3) BAC, the model achieved an AUC of 0.87, 95% CI = (0.72, 1.00), sensitivity of 0.80, and specificity of 0.93.

Conclusion: The framework holds promise for clinical adoption, integrating into mammography workflows and improving women's cardiovascular risk stratification.

Invasive lobular carcinoma (ILC) poses distinct diagnostic challenges due to its infiltrative single-file growth pattern, which often renders it mammographically occult, particularly in dense breast tissue. Contrast-enhanced mammography (CEM) combines the anatomical detail of conventional mammography with functional information from contrast uptake, likely improving the detection, staging, and assessment of ILC compared to conventional imaging techniques. CEM shows value in evaluating ILC tumor size and disease extent, especially in multifocal and multicentric disease, although MRI remains the gold standard. This review outlines the spectrum of ILC imaging features on CEM, including findings on both low-energy and recombined images. While CEM can provide ILC size and extent estimates comparable to MRI, its accuracy may be reduced in cases of non-mass enhancement or tumors larger than 3 cm. Additionally, ILC may demonstrate lower conspicuity enhancement than invasive ductal carcinoma (IDC), necessitating careful image interpretation. As clinical adoption of CEM increases, radiologists must become familiar with the variable imaging characteristics of ILC, to facilitate more accurate interpretation. Improved recognition of these features has the potential to support more precise treatment planning and better patient outcomes.

Radiology is experiencing rapid and interconnected change, including rising imaging volumes, expanding access demands, and the introduction of artificial intelligence into daily practice. However, many radiologists have limited exposure to structured approaches for leading change in complex clinical environments. Change management research provides a practical vocabulary and set of concepts that can help radiology leaders design and sequence change more effectively. Organizational readiness encompassing cognitive, operational, trust, and resource dimensions is consistently associated with successful transitions. Classic frameworks such as Lewin's change stages, Kotter's 8-step model for mobilizing teams, the ADKAR model for individual adoption, and Armenakis' evidence-based change-messaging principles offer radiology-specific value when planning workflow adjustments, introducing new processes, or shaping departmental culture. Attention to workflow reality, early engagement of key groups, understanding human responses to change, appropriate pacing, particularly during leadership transitions, and clarity of communication further support sustainable change. Applying contemporary change management concepts can help radiology departments and leaders navigate evolving demands while maintaining coherence, stability, and high-quality patient care.

Background: MRI often requires general anesthesia in children, which carries risks, increases costs, and prolongs scan wait times.

Purpose: Our study aimed to evaluate whether virtual reality (VR) simulations could familiarize children with the MRI experience to enable awake scans without anesthesia. Secondary objectives included assessing child anxiety and determining whether movement during the simulation correlated with scan quality.

Materials and methods: In this prospective study, 18 participants underwent a 10-minute VR simulation of an MRI procedure presented as an avatar-led game before their head MRI scan. Child and caregiver anxiety surveys were completed before the simulation and after the MRI. The VR software recorded head motion during the simulation, which was correlated with MRI scan quality.

Results: All participants (n = 18) successfully completed an awake MRI after the simulation session, aiding clinical diagnoses. The average participant age was 5.0 years (±1.3 years). MRI quality assessments indicated 44.4% excellent, 27.8% high-acceptable, 22.2% acceptable, and 5.6% low-acceptable scan quality. No statistically significant changes in anxiety levels were observed. 94.1% of legal guardians reported the VR simulation was effective at preparing their child for the MRI scan.

Conclusion: VR sessions were associated with a significant improvement in caregiver perceptions and enabled successful completion of MRI scans without the need for sedation in all children initially considered to require anesthesia. While no statistically significant reduction in anxiety was observed, the intervention resulted in diagnostic-quality imaging with minimal motion artifacts, supporting its utility as a strategy to facilitate pediatric MRI without anesthesia.

Background: Seizures are common neurological events in children, with neuroimaging playing a crucial role in evaluating new-onset seizures. While magnetic resonance imaging (MRI) is often preferred over computed tomography (CT) for pediatric seizure imaging due to higher sensitivity and lack of ionizing radiation, practices regarding imaging protocols and sedation use vary. Currently, there are no published Canada-wide guidelines describing the practices for workup of pediatric seizures.

Methods: A cross-sectional survey was conducted among radiologists at 16 Canadian tertiary pediatric centers to assess neuroimaging practices for children with new-onset seizures. The survey explored the presence and content of dedicated seizure MRI protocols, sedation use, and strategies to facilitate non-sedated MRI.

Results: Fifteen centers (94%) responded. Only 2 (13%) reported using dedicated new-onset seizure protocols, while 10 (67%) used epilepsy-specific MRI protocols, and the others used different approaches, including variations of routine brain MRI. MRI sequences varied across institutions. Sedation use also varied, with a median sedation age range of 3 months to 6 years. Non-sedated MRI techniques such as feed-and-swaddle (93%) and natural sleep (27%) were commonly used for infants. Video goggles (67%) and child life specialist support (53%) were often used for older children. Only 2 institutions (13%) had fast MRI protocols, and virtual reality preparation was uncommon (13%).

Conclusion: Considerable variability exists in MRI protocols and sedation practices across Canadian pediatric centers evaluating new-onset seizures. Our findings emphasize the need for national consensus guidelines to standardize imaging protocols, reduce sedation use, and optimize care for pediatric seizure patients.