ASFNR clinical state-of-practice: For posterior circulation large vessel occlusion stroke, use noncontrast CT to exclude hemorrhage, CT angiography to identify occlusion, and MRI with DWI as reference standard. Imaging scores like PC-ASPECTS help select patients for thrombectomy.
The ACR Appropriateness Criteria 2026 update provides imaging guidelines for suspected or confirmed diffuse lung disease, covering initial evaluation, acute exacerbation, and surveillance.
New expert consensus framework for symptomatic B3 breast lesions: assess concordance, lower excision threshold for discordant/palpable lesions, conservative management for concordant non-atypical lesions, and risk-based surveillance for atypia.
ACR Appropriateness Criteria for indolent lung cancer: For patients with slow-growing subsolid nodules or atypical cysts where surveillance is chosen over intervention, imaging follow-up aims to detect changes that alter management, with modality selection guided by accuracy and…
This AJR paper outlines a framework for translating AI-driven opportunistic screening findings (e.g., on CT) into standardized clinical actions, proposing report structuring and care pathways rather than raw model performance metrics.
Review of prostate-specific membrane antigen (PSMA) PET radiopharmaceuticals—including 68Ga-PSMA-11 and 18F-DCFPyL—covers clinical indications, protocol considerations, and workflow logistics for prostate cancer imaging.
An AJR review proposes reframing acute mesenteric ischemia as 'intestinal stroke' to emphasize the need for rapid diagnosis and revascularization, mirroring the urgency of cerebral stroke protocols.
The Society of Interventional Radiology (SIR) has released new practice guidelines in AJR for interventional radiologists managing chronic pelvic pain in women, covering patient selection, procedural techniques, and outcomes.
Pelvic venous findings are common in asymptomatic patients; labeling them as syndromes (nutcracker, May-Thurner) on imaging alone can lead to unnecessary care. A descriptive reporting framework may enhance communication and reduce overtreatment.
AJR study finds artificial intelligence cannot compensate for structural barriers that delay cancer diagnosis, suggesting AI tools alone are insufficient to fix systemic workflow issues.
New AJR study analyzes out-of-pocket costs for advanced diagnostic imaging among Medicare Advantage beneficiaries; results and implications remain to be fully detailed.
What does verbal feedback in radiology look like, and what should it look like? An AJR study explores this question, but the abstract is not provided; specific findings are not reported in the source.
A commentary examines a multimodal deep learning model for screening degenerative TMJ disease using low-radiation x-rays. It questions clinical translation without external validation. (exact figures not reported)
40-keV virtual monoenergetic images with deep learning reconstruction (DLIR-H) maximized hypovascular liver lesion detectability in a phantom study, suggesting a 31.3% potential for iodinated contrast media dose reduction. Larger body habitus limited this margin.
In radiographically normal rheumatoid arthritis, AI-measured bilateral joint space asymmetry at PIP3 and MCP3 joints on baseline hand radiographs predicted 2-year joint space narrowing with AUCs of 0.74 and 0.71, outperforming same-time high-resolution peripheral quantitative CT…
In a computational study, a multi-fidelity Gaussian process surrogate estimated porcine respiratory compliance and resistance from CT-derived models with errors below 5% relative to high-fidelity simulations, achieving over 100,000-fold speedup. Neural networks showed poor gener…
A new DICOM viewer architecture preserves diagnostic context by serializing viewer state for offscreen re-rendering, producing pixel-identical images (0 hash mismatches across 50 frames) to feed vision-language models, with capture latency of 4.49-16.69 ms.
Externally tested AI classifies lung nodules with high accuracy, supporting generalizability for screening. This benchmark study underscores the need for external validation before clinical deployment. (Exact metrics not reported in source.)
In a multicenter cohort of 6637 ED dizziness patients, ML models predicted stroke/TIA/dissection/tumor with AUC up to 0.97 (sensitivity 97%, specificity 91%), with hypothetical CT use reductions of 53-85%. External validation needed.
Radiomics, extracting quantitative features from routine imaging, shows promise for real-time detection and prediction of cancer therapy-related cardiotoxicity. However, current evidence is limited by small, non-externally validated studies. A paradigm shift in cardio-oncology i…
A perspective paper maps advances toward fully autonomous robotic-assisted and microrobotic surgery for endovascular, endoluminal, laparoscopic, ophthalmic, and orthopedic procedures, identifying key technical enablers needed for clinical translation.
Retrospective: patient-specific physics-constrained network reconstructs 3D CBCT from 2 orthogonal X-rays (PSNR 48.02 dB, SSIM 0.988, MAE 24.89 HU) outperforming GAN & diffusion models in 15 GYN cancer patients. No external validation.
A publicly released spatiotemporal MRI atlas of the developing fetal brain (21–37 weeks) from 159 typically developing fetuses now includes 126 anatomical regions and a deep learning segmentation tool (FetalSEG) for automated analysis.
Neuroradiology must 'Image Greenly'—adopting appropriate imaging, reducing contrast, MRI energy, and travel emissions—to mitigate healthcare's climate impact, says new ASNR-ESNR white paper.
Post hoc AI denoising reduced noise in neurovascular photon-counting CTA (p<0.001) but did not improve subjective IQ or diagnostic confidence in 42 pts, suggesting a ceiling effect when baseline quality is already high.
For deep learning detection of brain metastases, registration and pre-normalization reduced detection; bias correction/denoising had no effect. Tumor-core labels and fewer MRI sequences improved performance. Including all labels decreased detection to 68.8% (p<0.001) vs tumor-co…
Narrative systematic review of 14 studies: dose reductions of 5–94% with low-kVp/mAs and iterative/deep-learning reconstruction in pediatric chest/abdominal CT, maintaining diagnostic quality (meta-analysis not feasible due to heterogeneity).
In 307 cognitively impaired patients, vascular pathology independently linked to larger basal ganglia perivascular spaces (PVS) (Cohen d=1.28), while amyloid-β burden linked to lobar PVS. Mixed pathology showed less-than-additive effects, suggesting spatially distinct glymphatic…
A multi-slice deep learning model using alternating masks and double-contrast (T1w/T2w) input reconstructed brain MR images from 30% k-space data with a PSNR of 44.4 and SSIM of 0.987. The U-net approach outperformed two generative adversarial network models (DAGAN, SARA-GAN) wi…
In 2023, global TB incidence declined 19.2% since 2015, but 9.11M new cases and 1.22M deaths occurred. MDR-TB accounted for 466K cases. Sub-Saharan Africa and South Asia bear highest mortality. Progress is fragile, with recent health funding cuts threatening gains.
Systematic MRI phenotyping in OPHN1 disorder reveals a consistent constellation of cerebellar hypoplasia/dysfoliation, brainstem anomalies including anterior mesencephalic cap dysplasia, caudate hypoplasia, and frontal ventriculomegaly, expanding the recognized phenotype.
In CREST-H, intracranial atherosclerotic disease was most prevalent in ICAs (69% right, 67% left). Significant pairwise associations linked contralateral ICA, PCA, ACA (p<0.001) and right ICA–ACA (φ=0.25). Left ICA–ACA was not significant. Retrospective TOF-MRA analysis.
In SSPE, ASL perfusion revealed basal ganglia hyperperfusion in 54.5% (vs 15.2% structural MRI); a 4-region CNR composite AUC=0.90, 100% specificity, correlated with disability severity but not duration. Retrospective, 33 patients.
Photon-counting detector CT myelography led to 66% greater odds of a higher diagnostic confidence score for cerebrospinal fluid-venous fistulas compared with energy-integrating detector CT (OR 1.66; 95% CI 1.17-2.36), in a retrospective study of 628 exams.
