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…
Medicare's NTAP policy for an AI stroke triage tool is tied to unequal hospital adoption, raising concerns about worsening healthcare disparities, per a new Neiman Health Policy Institute study.
The Improving Seniors’ Timely Access to Care Act advanced out of a House committee, seeking to streamline prior authorization for medical imaging in Medicare Advantage.
Nano-X Imaging reported a net loss of $14.3 million, up 8% from the prior year, and issued a 'going concern' warning, signaling substantial doubt about its ability to continue operations.
Knowing the CT appearances of hepatic lymphoma—focal masses, diffuse infiltration, periportal disease—can aid in differential diagnosis of liver lesions. This educational video from CTisus reviews key patterns.
A new JVIR study by ACR Neiman HPI group reports that interventional radiology procedures are increasingly performed by a smaller subset of radiologists (exact figure not reported), raising concerns about workforce and training.
Toddlers with mild congenital heart disease (CHD) showed cortical gray matter volume and thickness reductions similar to moderate-to-severe CHD in a cross-sectional MRI study of 109 CHD and 56 healthy children (p<0.001 corrected). Alterations correlated with poorer motor and ada…
AI models for detecting gallbladder and bile duct stones achieve high diagnostic accuracy in research settings, but most studies are small, single-center, and lack external validation (exact performance metrics not available).
A 3D deep learning model differentiated temporomandibular joint disorders (TMD) from healthy condyles on CBCT with a moderate F1-score of 0.65. Performance was higher for specific TMD subtypes, reaching 0.76–0.88. This retrospective study indicates potential but requires larger,…
In a retrospective reader study of 80 COVID-19 chest X-rays across 8 radiologists, baseline accuracy was high and stayed consistent regardless of AI display design: fewer than 1% of correct calls flipped to incorrect with any AI overlay, while ~1% improved. Complex heatmap+graph…
回顾性研究中，GPT-4o-mini直观模型仅用术前数据预测输尿管镜碎石术后SIRS/败血症，AUC 0.973，敏感度92.9%，特异度95.2%。所有LLM模型组合AUC均>0.95，需前瞻性验证。
ASL perfusion radiomics AUC 0.955 in validation for distinguishing radiation brain injury from glioma recurrence, comparable to combined conventional+DSC (0.960). Suggests non-contrast alternative. Single-center retrospective study needs validation.
Deep learning uncovers a pre-therapy cardiac phenotype that predicts cardiotoxicity, per a study in JACC: Cardiovascular Imaging. The approach may enable earlier risk stratification.
AI applied to DXA yields superior fracture prediction (AUC ~0.90) versus standard BMD or FRAX, enabling earlier osteoporosis detection, per a narrative review.
A new machine learning framework, MeLSI, learns which specific microbes drive differences between groups in microbiome community comparisons. On the DietSwap dataset, MeLSI was the only method to achieve significance at α=0.05, detecting diet-induced shifts missed by standard me…
AI algorithms applied to abdomen/lumbar-spine CT scans show excellent performance for opportunistic osteoporosis screening, with a pooled AUC of 0.931 (95% CI 0.926-0.936). Screening for spine fractures was also strong at AUC 0.863 (0.73-0.936). DXA served as the reference. Most…
A flipped classroom model for veterinary radiology improved mean post-test scores by 2.9 points (6.5 to 9.5 out of 12, p<.001). Students appreciated flexible online modules but reported early overload and unclear expectations, while teacher support was mixed. (exact CI not repor…
A bibliometric analysis maps global and UK research on AI in medical imaging from 2017–2025, emphasizing generative and foundation models. (Exact figures not in source.)
Sparse Data Maximum Dispersion Classifier (SDMDC) outperformed other sparse models in simulated high-dimensional, low-sample-size data; a longitudinal version (LSDMDC) maintained performance with many predictors, especially under autoregressive correlation. (Simulation only, not…
A review maps how non-invasive imaging can assess cancer hallmarks: directly via probes targeting biomolecules or indirectly through pathophysiology, now boosted by AI-based radiomics and deep learning. Some of these imaging biomarkers already guide surgical, radiation, and mole…
Signal-oscillation analysis added to T2* mapping improved myocardial iron detection AUC from 0.79 to 0.94 in-vivo (0.78 to 0.99 ex-vivo), reducing fat-related false positives.
A multimodal AI model fusing chest radiographs with lab and demographic data predicted severe community-acquired pneumonia in children with ROC-AUC 92.91% (vs 78.28% for chest radiograph alone), in a retrospective study of 3,964 patients.
ML model OracleScreen-LILRB4 achieved 28.5% hit rate (15-fold enrichment) in virtual screening for LILRB4 binders, yielding 16 nanomolar compounds; lead compounds restored anti-tumor immunity in patient-derived cancer co-cultures.
Portuguese consensus: Renal denervation with radiofrequency or ultrasound may be considered for uncontrolled hypertension despite optimal therapy, based on sham-controlled trials showing sustained BP reductions and a favorable safety profile. Rigorous patient selection within mu…
In a retrospective 9-patient series, CT-guided epidural blood patch provided immediate symptomatic relief in all (100%) and sustained short-term resolution in 8/9 (88.9%) for postoperative lumbar CSF leak. No complications.
Retrospective single-center study of 471 stroke thrombectomy patients: isolated post-procedural subarachnoid hemorrhage was not associated with worse discharge NIHSS (P=0.98) or 90-day functional dependence (OR 0.69, 95%CI 0.28-1.68).
StrokeSENS ASPECTS, an AI tool, improved sensitivity for early ischemic changes on NCCT by 9.7 pp (p<0.001) and inter-reader agreement (κ 0.32→0.61) in an 8-reader, 100-case retrospective MRMC study of MCA/ICA occlusion patients.
In 7,477 neuroradiology publications (2016–2025), female first authorship rose from 21.5% to 31.6% and last authorship from 15.2% to 21.8% (P<0.001). Overall, women held 24.8% of first and 17.3% of last author spots. Senior female authorship remains low despite progress.
A single-center study of 559 patients found a MELD score > 16 predicts post-TIPS liver failure, with an AUROC of 0.785 for predicting liver-insufficiency related death (grade 3 PTLF). The overall incidence of post-TIPS liver failure was 17%, with a 6% mortality rate.
In a single-center retrospective study of 1,177 paracenteses, limiting PPE to sterile gloves alone did not raise post-procedure infection at 7d (1.1% vs 0.9%, p=0.41) or 14d (1.5% vs 1.8%, p=0.65).
In silico, deep contrastive learning model alzCL boosted Alzheimer's disease-associated gene detection by 18-24% AUC versus prior methods, identifying cell-type-specific targets and potential drugs like selonsertib, but lacks clinical validation.
