The 2026 Chinese lung cancer guideline adds AI and opportunistic screening in low-dose CT, expands molecular testing, and updates treatment regimens for NSCLC, including perioperative, consolidation immunotherapy, and targeted therapy for EGFR, ALK, KRAS, HER2, and MET alteratio…
On 76 test cases, the AI framework achieved Cobb angle MAE 2.4°, Lenke curve type accuracy 0.895, and reduced review time from 8.2 to 1.5 min/case.
An AI model (EAGLE) analyzing noncontrast chest CT achieved 90.0% sensitivity for esophageal cancer and 98.5% specificity in external multicenter testing (11,466 patients), and 99.94% specificity in low-dose screening. Could enable opportunistic esophageal cancer screening.
PSAD remained below 0.11 ng/mL² across all ages in 1037 healthy German men, contrasting with continuously rising PSA. 95th percentile reference: PV 51.39 mL, PSA 2.79 ng/mL, PSAD 0.071 ng/mL².
RSNA 2026 Plenaries Highlight Radiology’s Central Role in Healthcare
<![CDATA[FDA Clears Software That Combines MRI Enhancement of PET Imaging with Automated Quantification of Amyloid and Tau PET]]>
AI-assisted mammography showed 12% fewer interval cancers (MASAI trial) and greater sensitivity without specificity loss. Two cleared AI lung tools reduced false positives by 18% (RevealAI-Lung) and achieved 93% sensitivity (eyonis LCS). AI could decrease reading workload by 44%.
Facebook Live: Case Conference September '26
From Fellow to President: 35 Years of RSNA
Ultrasound MinnieCast, Episode 17: Elastography holds promise toward expanded role
<![CDATA[Does kV Still Matter in Photon-Counting CT?]]>
Study of 1.1M Medicare patients: only 57% with new cognitive impairment received recommended brain imaging within a year; 66% for dementia vs 51% for MCI. MRI used in just 46% of scans, with lower rates in dual-eligible and Medicare Advantage groups.
FDA Publishes Final Order on AI CAD Petition
House Panel Reviews ACR Priorities on Medicare Reform
ACR Urges CMS Transparency in 2027 Physician Fee Rule
AI full-vessel segmentation of intracoronary optical coherence tomography (OCT) in 100 consecutive patients matched expert reads: calcified plaque kappa 0.88 (95% CI 0.84-0.92), lipid plaque kappa 0.68, near or better than interobserver variability.
Quantitative MRI for renal cell carcinoma grading: from feasibility towards reproducibility.
Letter to the Editor: Can TLG and Ki-67 identify patients with metastatic soft tissue sarcoma who benefit from surgery?
Letter to the Editor: ADC as a non-invasive surrogate for PD-L1 in cervical cancer-addressing potential confounding and the leap to immunotherapy selection.
Letter to the Editor: Concordant breast-density measurement does not guarantee concordant risk classification-a downstream caveat for planning supplemental screening.
Letter to the Editor: Comparison of three algorithms to measure breast density on mammograms in a population-based screening cohort.
Radiomics and AI transform cancer imaging into quantitative phenotypic data that correlates with tumor biology. This review surveys classical and deep learning approaches and highlights the need for robust, reproducible methods before clinical adoption.
A deep learning model on rotational MIP FDG-PET achieved AUC 0.82 (95% CI 0.57–0.94) and 92% sensitivity for lymph node metastasis in esophageal squamous cell carcinoma, vs radiologists’ 58% sensitivity.
Hybrid/remote work aids recruitment amid a pediatric radiologist shortage, but groups must prioritize fairness, frequent multidirectional communication, and avoid overemphasis on individual wRVUs to prevent erosion of informal mentorship and trainee education.
An AI pipeline automatically analysed pre- and post-EVAR CTAs in 372 patients, segmenting anatomy with Dice >0.9, detecting endoleaks with AUC 0.85 on external validation, and measuring AAA diameter within 2 mm error (R≥0.97), matching interobserver variability.
Interpretable machine learning ensemble predicted 2-year hepatic encephalopathy after TIPS with AUC 0.960 (external 0.914) and variceal rebleeding with AUC 0.966, outperforming conventional scores (AUC 0.528–0.563). Top predictors: diabetes and recent bleeding.
In a retrospective study of 13 hemodialysis patients, the stepwise incision-free lumen-calibrated ligation (SILL) technique reduced mean access flow from 2516 to 921 mL/min (P<0.001) with technical success in all patients and clinical success in 84.6%, no complications.
ESUR Prostate MRI Working Group report: AI tools for prostate cancer detection on MRI show high technical accuracy but lack real-world validation. Key recommendations: human-in-the-loop use, prospective multi-vendor validation, and post-market surveillance.
A PACS-integrated AI tool for CT-based cervical lymph node assessment in papillary thyroid carcinoma achieved AUCs of 0.948 (internal) and 0.943 (external) for metastasis prediction, and significantly improved diagnostic accuracy among junior radiologists (p<0.05).
Multimodal model combining CT enterography radiomics and C-reactive protein (CRP) achieved AUCs 0.860 (internal) and 0.842 (external) for mucosal healing in small bowel Crohn’s disease, outperforming CRP alone (p<0.05) with prospective validation.
A knowledge-guided dual-path AI framework achieved macro-average F1-scores of 97.63% (internal) and 96.76% (external) for classifying liver MRI series across 18 categories, significantly outperforming an image-only model (p<0.001).
In 490 young adults, paternal education was associated with BMI and depressive scores, statistically mediated by structural-functional connectivity coupling in the right inferior parietal lobule (indirect effect 0.01, 95% CI 0.001–0.04). No maternal link found. Cross-sectional M…
Pre-treatment CT-based extranodal extension (iENE) did not predict recurrence-free survival in HPV-associated head and neck cancer. Gene signatures of inflammation captured prognostic information, suggesting tumor inflammation confounds iENE assessment (p<0.0001).
