Deep learning framework sharpens low-dose CT images while cutting noise
Researchers report a deep learning framework called ShearFocusNet that denoises low-dose CT images by combining Shearlet-based directional perception with spatial ...
Tag: EJNMMI Physics
Deep Learning Boosts PET Kinetic Parameter Estimation for Liver Cancer
A deep neural network-based trust region optimization algorithm outperformed conventional fitting methods in estimating kinetic parameters of hepatocellular carcinoma from ...
MRI-Based Centiloid Methods Prove Most Reliable for Amyloid PET Scoring
A head-to-head validation in EJNMMI Physics finds MRI-based Centiloid pipelines deliver near-perfect reproducibility for amyloid PET, while PET-only software performance ...
Deep learning sharpens ultra-low-dose PET/MR scans in Alzheimer’s disease feasibility study
A retrospective feasibility study reports that a deep learning network substantially improved simulated ultra-low-dose multi-tracer PET/MR images in Alzheimer's disease, ...
Ring-type semiconductor SPECT shows promise for dynamic imaging in phantom tests
Phantom experiments show a ring-type semiconductor SPECT/CT system can separate adjacent radioactive targets that overlap on planar imaging, though its ...
Alpha Versus Beta: Tumor Architecture Decides Where Radiation Dose Lands
Monte Carlo simulations show astatine-211 delivers about 120-fold higher single-cell nuclear dose than lutetium-177, but its effectiveness depends strongly on ...
New AI Model Turns Short Amyloid PET Scans into Full Late-Phase Images
A new generative AI framework called MOVIE synthesizes late-phase amyloid PET images from one-minute early-phase scans, outperforming a UNet baseline ...
POPULAR NEWS
-
Microbial Sources and Stress Shape the Hidden Endophytes of an Invasive Plant
-
Firebreaks Leave Lasting Scars in Grasslands While Wildfire Heals Fast
-
ZNF207 shields glutamine transporter ASCT2 from degradation to fuel liver cancer growth
-
Systematic review maps sustainability tensions in offshore wind development