Top 10 Stories from RSNA 2025 'Daily Bulletin'
Stories capture highlights from the most important week in radiology
Each year, Daily Bulletin highlights the discoveries, debates and innovations that define the RSNA annual meeting. This year offered no shortage of standout moments, including transformative applications of generative AI, new imaging approaches and a focus on workforce well-being.
We uncovered emerging insights in screening and diagnosis and shared compelling noninterpretive topics relevant to all medical imaging professionals. Our stories reflect radiology’s fast-shifting landscape. Here’s a roundup of the most-read coverage from the official newspaper of the RSNA annual meeting.
Why Everyone Is Having the Wrong Nightmares About AI
Humans often misjudge the long-term impact of transformative technologies, and AI is no exception, according to insights shared during a Tuesday plenary session. Techno-sociologist Zeynep Tufekci, PhD, MA, urging a shift from fear-based predictions to thoughtful planning for responsible deployment.
“When we discuss AI, we’re again using the wrong benchmarks and are ignoring the consequences of scale,” Dr. Tufekci said.
Emergency Radiologists Confront Burnout
Emergency radiology is uniquely demanding. High acuity and high volume converge with rapid turnaround expectations. Frequent interruptions and the need for 24/7 coverage create conditions that when sustained, create fertile ground for burnout, according to insights shared by Hemang Kotecha, DO, an emergency radiologist and assistant professor at UMass Memorial Health.
Burnout in emergency radiology often manifests as emotional exhaustion, depersonalization and a diminished sense of personal accomplishment, he explained.
“Burnout isn’t just an individual problem; it’s shaped by the environment and organizational structure,” Dr. Kotecha said.
Mammography To Do Double Duty
A Wednesday session highlighted how mammography can serve a dual role: screening for breast cancer and assessing cardiovascular risk in women.
Matthew Nudy, MD, of Penn State and Alyssa Watanabe, MD, of the University of Southern California Keck School of Medicine, showed that breast arterial calcification on mammograms, particularly when it progresses over time, is a strong predictor of major adverse cardiovascular events, offering a potential tool to identify high-risk patients early.
“A major conclusion of our study is that women who have progression of BAC on sequential mammograms are at an elevated risk of developing cardiovascular disease compared to women with stable or no BAC,” Dr. Nudy said.
Beyond Burnout: Rebuilding Radiologist Well-Being
Radiologist well-being remains a pressing concern, with burnout, isolation and moral injury affecting productivity and satisfaction, according to a Monday plenary session. Stanford Medicine’s chief wellness officer, Tait Shanafelt, MD, emphasized that while individual strategies help, meaningful change requires systemic solutions and leadership-driven cultural shifts in radiology departments.
“Twenty percent of the issues can be addressed by individuals, but 80% of the issues require system changes,” Dr. Shanafelt said.
Few Vertigo Cases Need Urgent Brain Imaging
Neuroimaging for emergency department (ED) patients with vertigo rarely uncovers acute, actional causes, with less than 2% of scans revealing serious central pathology, according to a 10-year study. Careful assessment of vertigo symptom quality, acuity and physical exam findings can help clinicians target imaging to patients most at risk, reducing unnecessary scans and associated costs.
“Our study showed, very importantly, that vertigo symptom quality, acuity and physical exam findings offer meaningful predictive value to ED physicians deciding which patients need urgent imaging,” said Raven Spencer, MD, third-year radiology resident at Allegheny Health Network in Pittsburgh.
AI Spots Frailty and Morality Risk from a Single Chest X-Ray
A new deep learning AI model can identify frailty from routine chest X-rays, offering a powerful tool to flag high-risk patients and potentially improve outcomes. The model demonstrated a strong association with mortality risk and biological age, highlighting AI’s potential to transform routine imaging into a preventative screening opportunity for older adults. n
“This research is a prime example of how AI can help unlock the hidden value in medical data we are already collecting, shifting patient care from reactive to preventative,” said Namkug Kim, PhD, of the University of Ulsan College of Medicine in Korea.
AI-Assisted Chest CTs Spot Heart Risks Early
An AI model can estimate the risk of major adverse cardiovascular events from routine non-contrast chest CT scans, offering a more accurate and equitable approach than traditional clinical risk models. Amara Tariq, PhD, highlighted that incorporating causal intervention allows the model to generalize across diverse patient populations, potentially enabling earlier detection and intervention for cardiovascular disease.
“Our model is not only more accurate than comparative models but is also more equitable and fairer,” said Dr. Tariq, a data science analyst at Mayo Clinic in Scottsdale, AZ.
Ultra-Low-Dose CT Shows Promise as First-Line Imaging for Trauma
A multicenter study across five French hospitals demonstrates that ultra-low-dose (ULD) CT scans detect fractures with accuracy approaching standard CT, offering a potential new first-line imaging option for trauma patients.
“This research confirms that ultra-low-dose CT is a compelling alternative to standard radiography for trauma patients. It offers high diagnostic confidence and near-gold-standard performance, while maintaining low radiation exposure which can enhance patient care and optimize department workflow,” said Oumaya Ben Said, MD.
Evaluating Generative AI-Assisted Radiology Reports with Clinical Precision
As multimodal generative AI models advance toward producing full draft radiology reports directly from imaging data, researchers are emphasizing the urgent need for clinically grounded ways to measure their accuracy. This education exhibit showcased a robust framework of evaluation metrics designed to ensure AI-generated reports are safe, reliable and ready for real-world use.
“By recognizing the gaps in current assessment methods, radiologists can more confidently integrate AI tools into their workflows—enhancing productivity without compromising diagnostic quality,” said Samantha Leech, PhD, an MD candidate at the University of Saskatchewan College of Medicine.
MicroUS an Effective Alternative to MRI for Prostate Cancer Diagnosis
Microultrasonography (micro-US)-guided biopsy is emerging as a fast, accurate and cost-effective alternative to multiparametric MRI with conventional US fusion for detecting clinically significant prostate cancer, according to results from the OPTIMUM Trial—the largest study of its kind to date.
The trial demonstrated that microUS is noninferior in detecting higher-grade cancers while offering real-time, high-resolution imaging that can streamline diagnosis and improve patient care.
“MicroUS imaging is a useful tool for both risk stratification and biopsy in cases where mpMRI is logistically challenging, or where patient factors make avoiding the extra procedure preferable,” said Sangeet Ghai, MD, director of research and professor, Joint Department of Medical Imaging at the University of Toronto.Whether you attended in person or followed along from afar, you can continue exploring RSNA 2025’s most impactful science and education. Meeting sessions, scientific presentations and educational content remain available on demand through Virtual Access until Feb. 27, 2026.
Revisit your favorite sessions, catch up on what you missed, and dive deeper into the innovations shaping radiology’s future.
For More Information
Continue learning with Virtual Access—visit Meeting Central or use the meeting app.
Browse the Exhibitor Directory - explore our full exhibitor list, watch 200+ industry presentations, get product information and connect with industry representatives.
Access the self-service Evaluation & Credit Site to select and evaluate sessions.