Radiology

Trials, guidelines and clinical evidence across Radiology.

AI report generation model improves radiology accuracy and generalization

UNIRG-CXR used reinforcement learning to improve chest X-ray report quality, cut clinically significant errors, and generalize across datasets.

A Dutch retrospective study found GPT-5 and other models could label most incidental CT findings in lung cancer…

A saliency-guided MRI framework improved brain tumor classification accuracy and robustness in public datasets, with Bayesian confidence estimation.

A deep learning pipeline segmented the cranio-cervical venous system on contrast MRI, sharply reducing manual annotation time and…

NGU-Net adds a feature refinement block to U-Net, improving brain MRI segmentation efficiency and accuracy in tumor-focused imaging…

MER-CycleGAN improved unpaired brain MRI synthesis and scanner harmonization, with better reconstruction metrics and modest downstream gains.

Grimsby hospital adds a third MRI scanner in a £2m expansion, raising daily capacity and improving access for…

A strict patient-level study found several CNNs generalized well for odontoid fracture detection, while others collapsed without hidden…

A quantum-inspired optimization framework improved medical image watermarking quality and robustness in chest X-ray and diabetic retinopathy tests.

St. Jude is hiring a postdoctoral fellow to build machine learning and foundation models for medical imaging, radiology,…

Radiology

Trials, guidelines and clinical evidence across Radiology.

In Brief The 5-minute rounds