Radiology Advances Podcast | RSNA

This episode explores a study from The First Affiliated Hospital of Anhui Medical University in Hefei, China, with Mayo Clinic collaborators, testing whether cardiac MR elastography can quantify left ventricular stiffness reliably enough for longitudinal use. Sixty-nine participants — 60 healthy volunteers and 9 patients with aortic stenosis or long-standing hypertension — underwent baseline scans, five-minute rescans, and one-month rescans, read independently by two radiologists. Agreement was excellent across readers and intervals, with stiffness ICCs of 0.96 to 0.97, and the study defined the smallest change that counts as real — roughly 0.6 to 0.9 kilopascals — paving the way for tracking therapies that soften the heart.

Reproducibility of 3.0 T single-frequency cardiac MR elastography for quantification of left ventricular myocardial stiffness. Wang. Radiology Advances, 2026, 3, umag032

A study from Radboud University Medical Center in the Netherlands compares unassisted, AI-assisted, and expert-assisted RECIST measurements on follow-up body CT's. 23 readers from 11 institutions measured target lesions in 212 cancer patients under all three conditions. AI assistance cut reading time by about a third, from 105 to 71 seconds per patient, and improved inter-reader agreement RECIST outcomes agreement by 7.7 percent, though expert double-reading remained even more consistent. The findings position AI as a scalable reader-in-the-loop tool for treatment response assessment.

Multicenter AI- versus Expert-Assisted RECIST Target Lesion Measurements in Follow-Up Body CT of Cancer Patients. Grauw et al. Radiology Advances, umag028.

A study from the Technical University of Munich in Germany comparing an AI-accelerated 3D gradient echo MRI sequence (CSAI-GRE) with ultrashort echo time MRI for lung nodule detection, using CT as the reference standard. In 54 patients with 97 nodules, CSAI-GRE detected 96.9% of nodules with higher image quality than UTE. But management-relevant Lung-RADS reclassifications relative to CT occurred in 5.6%, suggesting MRI-based nodule management is not yet ready for routine clinical use.

AI-accelerated 3D gradient echo versus ultrashort echo time MRI for lung nodule detection and measurement. Marka et al. Radiology Advances, 2026, 3, umag029.

This episode explores a study from Seoul National University Bundang Hospital in South Korea on opportunistic osteoporosis screening, which reads bone density off the spine in routine chest CT scans. Using nearly six hundred patients each scanned at two different tube voltages, the team built a model that converts those readings back to a single common reference, regardless of scanner settings or spinal level. The converted diagnostic thresholds closely matched directly measured ones, pointing toward more consistent screening from scans patients already have.

Standardizing Attenuation across Tube Voltages and Vertebral Levels for Opportunistic Osteoporosis Screening on Low-dose Chest CT. Kim et al. Radiology Advances, Volume 3, Issue 3, May 2026, umag026.

This episode discusses a study from Johns Hopkins University in the United States that introduces a brain age predictor from MRI, designed to stay accurate across different scanners and cohorts. The model estimates a person's age from a brain scan to within about four years, and holds that accuracy on an independent external dataset. The gap between predicted and actual age rose steadily from healthy adults to mild cognitive impairment to dementia, and tracked cognitive test scores, supporting its use as a marker of accelerated brain aging.

OpenMAP-BrainAge: Generalizable and Interpretable Brain Age Predictor from MRI. Kan et al. Radiology Advances, umag025

This episode covers a study from the Mallinckrodt Institute of Radiology at Washington University in St. Louis evaluating whether patient-specific lung shunt fraction measurement is necessary for every yttrium-90 selective internal radiation therapy case. Across 354 cases, the authors propose a new pretreatment metric called LSFbound — derived from liver mass, lung mass, and dose thresholds — that identifies which patients can safely skip the macroaggregated albumin imaging workflow. For most cases without large tumors or macrovascular invasion, the lung shunt simply does not constrain treatment planning, offering a path to a streamlined, single-session workflow.

When is patient-specific lung shunt fraction necessary in 90Y selective internal radiation therapy of liver cancer? Thomas et al. Radiology Advances, 2026, 3, umag007.

This episode discusses a study from New York University evaluating whether deep learning can predict acute pancreatitis severity from contrast-enhanced CT acquired within 24 hours of admission. Using self-supervised pretraining on about 12,000 unlabeled scans followed by supervised fine-tuning, the model achieved an AUROC near 0.89 for severe pancreatitis on both an internal NYU test set and an external multicenter Hungarian cohort of 518 patients, outperforming traditional clinical and imaging-based scoring systems. The work suggests that opportunistic AI triage on routinely acquired CT could support earlier, more accurate risk stratification in the emergency department.

Deep learning-based prediction of acute pancreatitis severity from abdominal CT with multicenter external validation. Xu et al. Radiology Advances, 2026, 3, umag020

This episode discusses a study from the University of California, San Francisco in the United States that tested whether GPT-4o-generated patient-friendly summaries improve comprehension of lung cancer screening CT reports. In a within-subjects survey of 1,815 adults across Lung-RADS 1, 2S, and 4B vignettes, the summaries significantly improved objective comprehension and reduced anxiety for all three report types. Largest gains were in participants with low self-rated English and health literacy. These findings support using LLM summariesas a potential health-equity tool, while highlighting the unmet patient need for personalized next-steps guidance.

Self-reported comprehension of large language model-generated summaries of lung cancer screening reports: a vignette survey. Serna et al. Radiology Advances, 2026, 3, umag008.

This episode explores a study from the University of Texas Southwestern Medical Center and MD Anderson Cancer Center in the United States that clinically validates an FDA-cleared AI tool for measuring total cardiac volume on non-contrast, non-gated chest CT. Across 307 patients with paired echocardiography, the AI discriminated normal from abnormal cardiac volume with an AUC of 0.81 in men and 0.77 in women, and far outperformed routine radiologist sensitivity for cardiomegaly. The tool offers a tunable, reproducible opportunistic screening layer on chest CT's already being performed.

Radiology Advances, 2026, 3, umag013. Fan et al.

This episode explores a study from the Emory Sports Performance and Research Center and the University of Lausanne that determined how few annotated MRI exams are needed to train a reliable deep learning model for thigh muscle segmentation. Using the nnU-Net framework with incrementally larger training sets, the researchers found that just 20 high-quality annotated subjects produced clinically acceptable segmentation across 14 thigh muscles, with biomarker agreement virtually indistinguishable from expert manual segmentation. All tools and trained models have been made openly available.

Optimizing MRI annotation workflows for high-accuracy deep learning thigh muscle segmentation in athletes. Slutsky-Ganesh et al. Radiology Advances, 2026, 3, umag005