Deep Learning Model Detects Simulated Tooth Ankylosis in Phantom CBCT Study
A deep learning model utilizing the YOLOv8 architecture successfully identified simulated tooth ankylosis on cone-beam computed tomography scans with high ...
A deep learning model utilizing the YOLOv8 architecture successfully identified simulated tooth ankylosis on cone-beam computed tomography scans with high ...
A controlled empirical study demonstrates that training an adapted U-Net on localized 256 by 256 patches increases microaneurysm pixel density ...
A deep-learning model trained on ultra-widefield fundus images accurately segmented myopic atrophic lesions, with performance gains tapering beyond 300 to ...
SCAD-Net improves seismic fault identification accuracy and spatial continuity across differing surveys using an adaptive domain adversarial architecture.
A retrospective study shows that quantified baseline ascites volume above 613.5 milliliters independently predicts significantly reduced survival in patients receiving ...
Evaluating lung adenocarcinoma whole-slide images revealed that mild JPEG2000 compression reduces storage requirements by approximately 90% while maintaining deep learning ...
Deep learning models trained on sparse expert annotations can segment anatomical structures in pediatric swallowing endoscopy videos from unseen patients, ...
A SegFormer-B2-based multi-expert ensemble called MUSE-Net achieved a mean Dice of 0.834 for carotid plaque segmentation across a four-center ultrasound ...
A Chinese research team reports a new asymmetric U-Net with hybrid attention and residual dilated convolution that achieved Dice scores ...
A new MLP-based segmentation method combining mixer layers with channel attention achieved a Dice coefficient of 81.45 percent on the ...
© 2026 Scienmag - Science Magazine
© 2026 Scienmag - Science Magazine