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.
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.
Noncontrast CT-based unimodal imaging AI detected anterior circulation large vessel occlusion with sensitivity 0.78 (95% credible interval 0.68-0.86) and specificity 0.88, versus expert readers' sensitivity 0.62 and similar specificity; low-certainty retrospective evidence.
SIR Foundation research consensus panel calls for standardized acute DVT intervention endpoints: biologically informed patient selection, anatomic/physiologic metrics, validated severity tools, longitudinal biomarkers, and economic/QoL measures to advance precision and reduce po…
In 200,000 screening exams, an AI model's AUC for breast cancer detection was highest in the least dense breasts (0.955) and lowest in the densest (0.857). Density distribution varied by vendor, and 18.5% of exams had different density categories for the right and left breast.
Letter to the Editor: Universal landmark matching in musculoskeletal radiography: flexibility should be matched by reliability safeguards.
Reply to the Letter to the Editor: Sequential dual-energy CT for longitudinal assessment of pathologic response to neoadjuvant immuno-chemotherapy in locally advanced gastric cancer.
Letter to the Editor: Statistical reporting and model transparency in sequential dual-energy CT assessment of treatment response in gastric cancer.
In 1847 breast MRI exams, a DINOv2-based medical slice transformer triaged cases without suspicious findings (BI-RADS ≥4) at 97.5% sensitivity. Specificity was 19% for contrast-enhanced (T1sub+T2w) and 17% for non-contrast-enhanced (DWI1500+T2w) MRI. External validation AUC 0.77.
AI in pediatric imaging lags behind adult radiology AI due to developmental variability, differing disease epidemiology, and limited data; narrative review identifies key barriers and opportunities for safe clinical translation.
Artificial intelligence (AI) in breast imaging has moved beyond retrospective accuracy studies to prospective randomized controlled trials, with roles in cancer detection, risk prediction, workflow optimization, and supplemental imaging.
A Tale of Two Countries: Challenges in the Adoption of Artificial Intelligence in Medical Imaging in India and Singapore
ChatGPT-5 outperformed ChatGPT-4o for detecting lumbar spondylolisthesis on standing lateral radiographs, with sensitivity 67.5% vs 49.4% and accuracy 61.5% vs 55.5%, but both remained limited compared with fellowship-trained spine surgeons.
Multiomics deep learning combining computed tomography (CT) radiomics, pathology features, and clinical variables predicted microsatellite instability in colorectal cancer with AUCs of 0.996, 0.999, and 0.993 in training, internal, and external validation (n=509, 2 centers).
FDG-PET/CT is central to lymphoma management: Lugano staging, Deauville response assessment, and total metabolic tumor volume as a powerful prognostic biomarker. In Hodgkin lymphoma, interim PET guides treatment de-escalation; in DLBCL, it is prognostic but not yet actionable. P…
AI post-processing of standard-dose contrast-enhanced brain MRI improved lesion conspicuity: contrast-to-noise ratio up 495.16%, lesion-to-brain ratio up 58.94%, contrast enhancement up 160.55%; readers preferred it in up to 84.9% of cases.
An editorial says multimodal MRI and artificial intelligence/machine learning improve differential diagnosis of Parkinson's disease versus atypical parkinsonian syndromes; clinical use needs standardized protocols and reporting.
Gemini-3 VLM achieved 90.3% accuracy (95% CI 81.0%-95.5%) for zero-shot detection of condylar osseous changes on CBCT, outperforming GPT-5.2 (75.0%) and Qwen3-VL (55.6%) in 72 internal and 70 external images.
Deep-learning denoising raised low-field 0.55T knee MRI sensitivity from 0.83 to 0.97 and diagnostic accuracy from 0.83 to 0.98, approaching conventional 3T (sensitivity 0.95, accuracy 0.97) in a retrospective study of 33 knee MRIs.
Contrast leaked onto CT detector created peripheral dark artifacts mimicking stroke on cranial exams. Confirmed via phantom; radiologists must recognize this scanner-based artifact for quality control and to avoid misdiagnosis.
Change in mean tumor diameter on noncontrast CT showed AUC 0.83 for major pathological response to neoadjuvant immunotherapy in resectable NSCLC, with 76% sensitivity and 84% specificity, and numerically outperformed RECIST 1.1 (79.2% vs 72.9% accuracy).
A random forest model using non-contrast CT radiomics predicted spontaneous stone passage in acute ureteric colic (AUC 0.79, 95% CI 0.68–0.88; n=428, spontaneous passage rate 47.9%). External validation needed.
A transformer-based model combining DCE-MRI habitat and conventional radiomics with clinical predictors achieved an AUC of 0.923 (95% CI 0.858-0.988) for predicting microvascular invasion in hepatocellular carcinoma, and stratified recurrence-free survival (p=0.033).
FDA cleared GE Medical Systems' AIR Recon DL, a deep learning MRI reconstruction software, under 510(k) (product code LNH).
In breast cancer survivors, AI-derived vertebral bone density (vBMD) and paraspinal muscle-fat ratio on routine CT helped identify moderate-severe vertebral fractures. AI-vBMD AUC 0.738, adding muscle-fat ratio improved AUC to 0.786, full model AUC 0.828.
A random forest model using PET/CT and clinical factors (heart rate variability, LDL cholesterol, age, monocytes) predicted ischemia in nonobstructive coronary artery disease (INOCA) with internal AUC 0.92 (95% CI 0.89-0.95) and external AUC 0.71 (95% CI 0.63-0.79).
NIH diagnostic radiology research funding rose from $39.9M in 2012 to $87.9M in 2025, but clinician-scientists held just under 13% of principal investigator roles in radiology since 2012, an Academic Radiology analysis found.
DEXA reveals tea consumption supports bone health
ASTRO Foundation awards $2.4M in radiation medicine grants
Fellowship trained neuroradiologist for partnership track or employee position at thriving physician-owned private practice affiliated with Yale Job Opening in Greenwich, Connecticut
James Clinic Employer Profile
AI tools may not meet breast radiologists' expectations in practice
Conavi Medical nets $10M in equity offering
Ventripoint expands access to cardiac imaging platform
