Deep Learning Framework Automates Lenke Classification on Multi-View Spine Radiographs
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.
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…
Two Clinical Scenarios, One Responsibility: The Radiologist Across the MASLD Spectrum.
Better, Stronger, Faster: Machine Learning and the Next Era of Accelerated MRI.
Pancreatic Cystic Lesions and the Need for a Paradigm Shift.
MRI Standards for Prostate Cancer Screening in the PRISM Consensus Recommendations.
Finding the Right Balance After a Negative PSMA PET/CT Examination.
Zero-Shot Performance of Six Vision-Language Models for Lung Nodule Detection on Chest Radiographs.
Biopsy Versus Clear-Cell Likelihood Score for Diagnosing Clear Cell Renal Cell Carcinoma in Small Renal Masses on MRI: A Propensity Score-Matched Analysis.
Beyond the Prostate Feeding Artery: Biologic Validation and Future Directions.
Artificial Intelligence in Medical Imaging: Navigating the Artificial Intelligence Trilemma Toward Responsible Clinical Translation
In rural Nigeria, 33% of 9,585 AI-screened chest X-rays were abnormal. TB was confirmed in 204 people (95% started treatment), but among the 2,367 with suspected chronic respiratory or cardiovascular disease, only 12% completed the referral.
A new commentary proposes a six-part framework and 90-day roadmap to shift radiology AI from fragmented point solutions to a managed operational capability with shared governance, integration, and measurement.
Combined radiomics-clinical model from baseline CT predicted PFS in HR+/HER2- advanced breast cancer on CDK4/6 inhibitor + endocrine therapy (C-index 0.85 training, 0.76 testing) but requires external validation.
Combining T1 contrast-enhanced and T2-weighted MRI radiomics with habitat analysis achieved AUC 0.863 (95% CI 0.763-0.954) for meningioma subtyping, outperforming radiomics-only and habitat-only models.
Advances in detector technology, quantitative PET/SPECT, patient-specific dosimetry, and AI are transforming pediatric nuclear medicine from standardized activity administration toward personalized radiation dose optimization, maximizing diagnostic and therapeutic benefit while…
Synthetic contrast-enhanced CT from a GAN improved gastric cancer histologic grade prediction AUC by 5.9% (internal) and 0.7% (external) over non-contrast CT in a 1,757-patient multicenter study. Lauren classification prediction improved 6.8% and 7.3%.
A deep learning algorithm for MRI-based abdominal body composition achieved Dice scores of 0.862-0.923, and strong correlation (r>0.95) with CT, but wide 95% limits of agreement for visceral fat (-8.0% to 35.9%) indicate need for modality-specific norms.
Bone-RADS 1 on MRI had 100% NPV for vertebral lesions (95% CI 94.7-100%, n=46). Bone-RADS 2/3 lesions had 3.4% malignancy rate, falling to 0% with clinical criteria. Inter-reader agreement was moderate (κ=0.60).
ASL-derived fractional tumor burden (FTB)-high differentiated recurrent tumor from radiation necrosis in 102 brain metastases after SRS (AUC 0.81, sensitivity 0.79, specificity 0.74). It performed comparably to perfusion metrics and outperformed delta T1.
In 90 CT-guided lung cryoablation procedures, a higher trajectory-based trauma index (TTI) strongly associated with pneumothorax (OR 21.87, P<.001) while greater pleural infiltration reduced odds (OR 0.56, P<.001). 62% had pneumothorax.
ESR Bridges: chronic thromboembolic pulmonary hypertension: new developments in imaging and treatment-a multidisciplinary view.
In 73 metastatic melanoma patients on immunotherapy, baseline peripheral rim iodine concentration on DECT independently predicted shorter progression-free survival and greater new-lesion burden, but not individual lesion progression. Organ-specific effects were heterogeneous.
At 13±2 s after exercise, biventricular volumes matched peak effort (all P=1.00), then declined linearly; left ventricular stroke volume fell 2.1 mL per 10 s. Rapid post-exercise cardiac MRI is a practical surrogate with better image quality.
Deep learning computed tomography (CT) volumetry distinguished idiopathic normal pressure hydrocephalus (iNPH) from controls with area under the curve 0.97 (95% confidence interval 0.94–1.00) externally and 0.99 (0.98–1.00) in Uppsala.
Editorial: Emerging fast medical imaging techniques in radiology.
A multimodal model combining post-thrombectomy CT hyperdensity features and NLP-clinical embeddings predicted malignant cerebral edema (midline shift ≥5 mm) with AUC 0.800 (95% CI 0.700-0.901) on external validation, improving senior neuroradiologists’ specificity from 78.1% to…
Adjusted median CT-to-groin time fell from 47 to 35 mins and CT-to-perfusion map time from 7 to 6 mins after AI tool implementation; benefit concentrated at primary stroke centres; clinical outcomes unchanged.
ASTRO Members Elect Dr. Beth Erickson as President-Elect and Board of Director Members
LLM-generated clinical histories for imaging orders were more comprehensive and factual than those written by referring clinicians. Claude 3.5 Sonnet was rated most useful for protocoling and interpretation, per a Radiology study of >28,000 records.
DiffuDose AI model generates patient-specific radiation dose maps for radiopharmaceutical therapy in <23 seconds per patient, with accuracy comparable to the gold standard. It outperformed six competing approaches and accurately assessed dose in kidneys and liver.
New Data Highlights Potential for Focused Ultrasound in Prostate Cancer Treatment