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…
A scoping review found only 4 small studies on Pacinian corpuscle distribution in diabetic foot; both histology and MRI can detect them, but evidence is too sparse to establish reference standards for clinical use.
A new German guideline for silicosis assessment emphasizes low-dose volumetric HRCT over chest X-ray due to poor correlation with symptoms and function, standardizing expert evaluations for occupational disease claims.
Multi-reader validation of a Gemini-based LLM pipeline for radiology board prep found 1.00% of flashcards had critical errors (95% CI 0.50–1.78%), exceeding the 0.3% safety threshold. Attending radiologists detected more errors than residents (OR 4.52). Automated feedback improv…
Microvascular decompression via an inferior ventrolateral brainstem approach achieved long-term cure in ~90% of hemifacial spasm patients, with one-third experiencing delayed relief (n>300).
No abstract or findings available for summarization.
While GPT-4o reliably identified errors in breast imaging drafts and gave helpful feedback, a comment highlights 3 methodological concerns: construct circularity, selection bias from Levenshtein filtering, and a drop in attending interrater agreement when GPT-4o joined the panel.
ED neuroimaging use rose across all ages and U.S. regions from 2016–2025, largest absolute increase in adults ≥65 (from 377.5 to 592.6 per 1,000 encounters). All trends p<.001. Retrospective Epic Cosmos data.
A multiparametric MRI model (ADC, rCBF, rCBV, rK2) predicted parenchymal hematoma after thrombectomy with AUC 0.88 and 77% accuracy in a retrospective, single-center study of 100 patients.
A letter in JVIR argues that contralateral suppression indices for primary aldosteronism should be replaced by outcome-based reference standards to improve patient selection for adrenalectomy.
Combined MRI radiomics+clinical model predicted VETC positivity in HCC ≤3cm with AUCs 0.885 (training), 0.867 (validation), 0.849 (external validation). Also stratified early recurrence risk (p<0.05).
A meta-analysis of 13 studies (N=3173) finds combined intratumoral-peritumoral MRI radiomics models achieve a pooled AUC of 0.83 for predicting microvascular invasion in hepatocellular carcinoma. The incremental benefit over intratumoral-only (AUC 0.78) or peritumoral-only (AUC…
SAM2CT model converts radiologists' arrows and line annotations into 3D CT segmentations, achieving Dice coefficients of 0.649 (arrow) and 0.757 (line). On clinical PACS data (n=60), 87% of generated segmentations were clinically acceptable. Zero-shot ED findings showed DSC 0.61…
Radiologists missed up to 11% of clinically significant errors in LLM patient communications; dual-reader review cut miss rate from 12.4% to 0.6%.
CODAvision, a GUI-based deep learning tool, enables customizable segmentation of medical images (histology, MRI, CT) without coding expertise; protocol and best practices now published.
Deep learning using diffusion MRI and light-sheet microscopy generated a whole-brain neuronal cell-type atlas of the adult mouse brain at 10 µm isotropic resolution, registered to the Allen Mouse Brain Common Coordinate Framework.
A model combining EfficientNetV2-Small and PPO achieved 95.9% AUC and 98.4% sensitivity for pediatric fracture detection on radiographs (n=1221, single center). Outperformed baseline CNNs. Prototype; needs prospective validation.
An AI-assisted ultrasound nomogram predicted axillary lymph node metastasis in breast cancer with AUC 0.81, sensitivity 77%, specificity 82% (internal validation), outperforming radiologist assessment. Retrospective, single-center, not externally validated.
An expert review cautions that AI-driven molecular diagnostics may amplify existing healthcare disparities unless demographic biases in training data and algorithms are proactively addressed (exact figures not reported in source).
Retrospective study: an MRI radiomics+clinical model predicted MammaPrint result in HR+/HER2- early breast cancer with AUC 0.86 (training), 0.76–0.79 (external), outperforming clinical-only (0.69,0.66,0.67) and radiomics-only (0.77,0.67,0.68).
An AI tool showed near-perfect agreement (ICC 0.97-0.99) with expert readers for Cobb angle measurements on 144 pediatric EOS spine radiographs, identifying ≥5° progression with 91% accuracy. Mean differences were negligible (0.2°-0.7°), supporting consistent, efficient scoliosi…
In a retrospective study of 1000 dental panoramic X-rays, a YOLO11 deep learning model estimated gonial angle with mean absolute error 2.37° and RMSE 2.58°, achieving pose F1-score 0.998.
ImprovedVertebroV5, a 3D U-Net optimized for small structures, segmented vertebrobasilar artery calcification on cone-beam CT with mean Dice 0.73 and small-object sensitivity 0.94 in 20 patients (retrospective, single-center). Significant gains over prior model.
Pooled sensitivity 85% and specificity 87% (AUC 0.92) for radiomics-based differentiation of benign vs malignant breast lesions on 2D ultrasound, but likely upper-bound estimate due to high risk of bias and heterogeneity, per meta-analysis of 27 studies (9,627 lesions).
Chest X-ray deep learning model predicted ICU admission within 24 h with AUC 0.78 (95% CI 0.74–0.81), outperforming Charlson (AUC improvement 0.16) and age-adjusted Charlson (0.12) comorbidity indices (p<0.001). Retrospective, single-site COVID-19 cohort; external validation nee…
Fully automated AI workflow for orbital fracture mirroring achieved mean Dice 0.936 and angular deviation 0.95° (fractured), 0.72° (uninjured) vs manual planning, in a multicenter study of 502 CT scans.
Nursing students reported VR provides low-anxiety environmental immersion for cognitive rehearsal, while traditional simulation develops interpersonal skills, dynamic decision-making, and empathy. A sequential integrated pathway (VR rehearsal→traditional scenario practice→guided…
XGBoost model classifies concurrent posterior circulation infarction in vertebrobasilar dolichoectasia (AUC 0.884, 95% CI 0.842-0.925) using imaging and clinical variables. Internal validation only; external and prospective validation required. A web calculator is available.
A novel POCUS curriculum in Tanzania improved EM residents' knowledge by 38.9% and OSCE scores by 40.7%, with scan accuracy up 3.3% and strong inter-rater agreement (kappa 0.83) in a 17-resident pre-post study.
ACR BI-RADS v2025 adds new descriptors like "lobulated" and "non-mass lesion," renames MRI kinetics to "early phase," and separates assessment from management. It formally integrates contrast-enhanced mammography and supports surgical excision as definitive treatment under a rev…
Three commercial multimodal LLMs showed differing zero-shot operating profiles on chest radiographs. GPT-5.4 had highest discrimination (AUC-ROC 0.836-0.883) but sensitivity as low as 0.043, while Gemini 2.5 Pro favored sensitivity (up to 0.973). None are suitable for autonomous…