Stanford Radiology Highlights from RSNA 2025 - Awards, Presidential Address, and Collaboration Celebrations

The AI Development and Evaluation (AIDE) Lab (authors Magdalini Paschali, PhD, Asad Aali, Arogya Koirala, David Larson, MD, MBA, and Akshay Chaudhari, PhD) received a Magna Cum Laude abstract award for their project Can AI Be Trusted to Summarize Radiology Reports for Patients? A Robustness Evaluation Using Adversarial Edits. The work evaluates the reliability of AI-generated patient-facing report summaries by testing model performance under adversarial conditions. The study highlights how even small alterations in report text can influence the quality of AI summaries. The findings underscore the need for continued development of safeguards to ensure accurate and trustworthy patient communication.

The Magna Cum Laude awards are the highest abstract honors handed out at the annual RSNA conference. During the entire meeting, only 3 Magna Cum Laude awards were presented among a total of 168 scientific posters shown.

The AIDE Lab (co-first authors Andrew Johnston, MD, MBA and Arogya Koirala, along with Lina Cheuy, PhD, Hye Sun Na, Jesse Sandberg, MD, Sergios Gatidis, MD, Akshay Chaudhari, PhD, and David B. Larson, MD, MBA) also received a second Magna Cum Laude recognition for their project Developing AI for Pediatric Radiology: Automated Leg Length Measurement Use Case. The exhibit presented specific considerations for pediatric populations when developing artificial intelligence tools for imaging. Their work emphasizes the importance of building pediatric-specific datasets and validation strategies to ensure safe and reliable clinical deployment.

This award represents one of 18 Magna Cum Laude awards presented among the 1134 education exhibits shown at the conference. 

Pejman Ghanouni MD, PhD, and Geoffrey Sonn, MD, received a Cum Laude award for their project CAPTAIN Randomized Controlled Trial of MRI-Guided Transurethral Ultrasound Ablation (TULSA) Versus Robotic Radical Prostatectomy: Initial Perioperative Outcomes. The study presents early perioperative findings comparing MRI-guided TULSA with robotic radical prostatectomy in the treatment of prostate disease. The project highlights the importance of evaluating emerging minimally invasive techniques in a controlled trial setting. It also demonstrates the value of early perioperative data in guiding future research and clinical practice.

This award represents one of 6 Cum Laude Science awards presented among the 249 scientific poster discussions shown at the conference.

Aya Kamaya, MD, and the project team received a Certificate of Merit for Sound Decisions: Ultrasound in Chronic Liver Disease Assessment and Surveillance. The work highlights the role of ultrasound in evaluating and monitoring chronic liver disease, with a focus on improving diagnostic accuracy and long-term patient surveillance. The project underscores the need for consistent imaging protocols to support reliable assessment across patient populations. It also reinforces the importance of integrating ultrasound findings into broader clinical pathways for chronic liver disease management.

This award represents one of 202 Certificate of Merit awards presented among the 113 education exhibits shown at the conference.

Andrew Johnston, MD, MBA, Sabine McHugh, Gloria Hwang, MD, Uzma Waheed, MD, and the project team received a Quality Improvement Reports Award for Supporting Innovation in Care Delivery: Reducing Turnaround Times for Same Day Imaging Exams Requested by the Oncology Rapid Assessment Center. The project focuses on improving imaging turnaround times to better support timely oncology care and streamline same-day assessment workflows. The team outlined process adjustments aimed at removing bottlenecks and coordinating care more efficiently. The initiative provides a framework that other services can adapt to enhance patient access and operational performance.

This award represents one of 6 Quality Improvement Reports awards presented among the 14 quality improvement reports shown at the conference.

On November 30, incoming chair Umar Mahmood, current president of the Radiological Society of North America (RSNA), delivered his presidential address at RSNA 2025, reflecting on the specialty’s evolution and its future trajectory. “Radiology today is a different specialty than it was 10 years ago. And because of our hard work together, it’s going to be a different specialty 10 years from now,” he noted.

Dr. Mahmood emphasized that radiology has always been about more than images, highlighting the field’s expanding role in patient care. Echoing the meeting theme Imaging the Individual, he underscored that while clinicians share common goals, each patient has unique needs and biology.

Stanford Radiology has been a longstanding collaborator with GE HealthCare, including on its photon counting CT (PCCT) technology, leading to the installation of a research PCCT scanner in Stanford Research Park in 3155 Porter Dr in October 2023.

At RSNA, the technology developed from this partnership was announced by GE HealthCare as the Photonova Spectra PCCT scanner (510k pending). At the launch event on November 30, 2025, Nancy Pham, MD (Neuroradiology) and Brian Pogatchnik, MD (Cardiovascular Imaging), shared their perspectives on how this new technology has the potential to provide new clinical insights due to its higher spatial resolution, rich spectral information, and potential for radiation dose reduction.

Researchers at Stanford Radiology are already using the installed research PCCT scanner to carry out both human-subject and technical studies to assess advanced image-reconstruction methods, maximize the scanner’s spatial and spectral capabilities, and uncover pathology- and disease-specific benefits enabled by the PCCT technology, such as the potential to develop new biomarkers for knee health.