30-Second Takeaway
- Gadoxetate-enhanced MRI markedly outperforms ultrasound in cirrhotic HCC surveillance image quality, especially for technically limited patients.
- CT-derived FFR, high-risk plaque features, and quantitative plaque burden refine CAD risk beyond stenosis alone.
- Super-resolution deep learning reconstruction improves CCTA diagnostic performance and frequently reclassifies CAD-RADS categories versus hybrid iterative reconstruction.
- Bone marrow MRI, preferably whole-body with DWI and Dixon, should replace radiographs and bone scans for many oncologic indications.
- Structured frameworks like AI-RADS will be increasingly necessary to document case-level AI reliability and radiologist overrides in reports.
Week ending February 14, 2026
Imaging quality, AI integration, and coronary CT risk markers: pragmatic updates for daily radiology practice
Gadoxetate MRI delivers far superior image quality to ultrasound for HCC surveillance in cirrhosis
In 245 cirrhotic patients undergoing paired HCC surveillance, gadoxetate-enhanced MRI showed no or minimal limitations in 80% versus 24% for ultrasound. Severely limited examinations were rare with MRI (1%) but common with ultrasound (14%), underscoring MRI’s technical robustness for surveillance. Obesity reduced image quality for both modalities, while Child-Pugh B/C cirrhosis independently worsened MRI quality on multivariable analysis. These data support gadoxetate MRI as a higher-yield surveillance option than ultrasound, particularly where US visualization is frequently suboptimal.
International consensus standardizes clinical use of CT-derived FFR in coronary CTA
This expert consensus summarizes evidence that CT-FFR improves detection of functionally significant CAD over anatomic CCTA alone, especially in intermediate lesions. Meta-analytic data show per-vessel accuracy roughly 71%–91%, with sensitivity 76%–98% and specificity 61%–94% against invasive FFR. The document advocates CT-FFR as a gatekeeper to invasive angiography and as a tool for PCI planning and prognostication. It also addresses technical and physiologic factors affecting accuracy and concludes that CT-FFR is cost-effective versus traditional noninvasive strategies.
CT characterization of incidental adnexal lesions is reliable for only a few entities
In a 75-patient, multireader study, CT readers most accurately diagnosed dermoids, malignant ovarian lesions with metastases, and simple cysts. These three entities showed high accuracy (up to 99%) and substantial-to-almost-perfect interreader agreement (Gwet AC1 up to 0.97). All other adnexal lesion types had accuracy below 72% and only fair-to-moderate agreement, highlighting substantial diagnostic uncertainty. Malignant ovarian lesions without metastases were misclassified as benign in 28% of interpretations, emphasizing the need for caution and follow-up imaging.
ESR recommends MRI—preferably whole-body—for bone marrow assessment in oncology
These European Society of Musculoskeletal Radiology recommendations position MRI as the primary modality for suspected bone metastases and myeloma staging. Whole-body MRI is preferred for comprehensive assessment, with axial-skeleton MRI as a faster alternative for targeted or follow-up evaluations. Standardized protocols should include fat-sensitive fast spin-echo T2 Dixon and diffusion-weighted imaging as fundamental components. Quantitative biomarkers such as ADC and fat fraction are recommended to improve diagnostic accuracy and response assessment across malignancies.
References
Numbered in order of appearance. Click any reference to view details.
Additional Reads
Optional additional studies from this edition.