Tag: Medical Imaging

A new narrative review finds AI tools improving diagnostic sensitivity and workflow efficiency in radiology while warning of demographic bias, ...

A phantom study found that a spectral photon-counting CT prototype, tuned to the STD-B kernel at iDose4 level 9, outperformed ...

A new Vision Mamba-based network with enhanced edge-guided attention achieves competitive optic disc and cup segmentation across four fundus image ...

A preliminary five-patient study found that a prototype deep silicon photon-counting CT scanner produced virtual unenhanced images within 10 Hounsfield ...

Researchers report a deep learning framework called ShearFocusNet that denoises low-dose CT images by combining Shearlet-based directional perception with spatial ...

A new hybrid CNN-Transformer model called AGDE-Net accurately segments tumor cores and peritumoral edema in adult glioma MRI while keeping ...

A new perspective article argues that coronary CT angiography, enhanced by artificial intelligence, could help clinicians stratify cardiovascular risk in ...

A hybrid CNN-Transformer framework called LHCT-Net reached 98.55% accuracy for melanoma-versus-non-melanoma classification on the 25,331-image ISIC 2019 dataset.

An explainable transformer model fusing MRI with texture descriptors reached over 99 percent accuracy in Alzheimer's disease staging benchmarks, with ...

An AI model trained on nearly 6,000 annotated uterine ultrasound images automatically segments the endometrium, measures its thickness with sub-millimeter ...

A meta-analysis of twenty studies found point-of-care lung ultrasound diagnosed childhood pneumonia with higher pooled sensitivity and specificity than chest ...

A selective dual-key encryption scheme uses hybrid chaotic maps to protect critical regions of medical CT images while supporting real-time ...

A new 3D Mamba-based model segments brain tumors with consistent accuracy gains and a 15x efficiency improvement over existing state-of-the-art ...

A systematic review of 26 studies finds AI models diagnosing multiple sclerosis with over 95 percent accuracy in some cases, ...

SalientNet is an unsupervised framework that generates mechanistically grounded saliency maps for CNN medical image classifiers, outperforming Grad-CAM, Grad-CAM++, and ...

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 pilot study reports that an infrared heating and imaging system distinguished cancer-free from cancer-positive surgical margins with 93.33% accuracy ...

A benchmark study couples CNN saliency maps, a cortical atlas and three LLMs to generate interpretable diagnostic reports from brain ...

A comparative study of nine U-Net-based deep learning configurations found U-Net with a VGG16 backbone achieved the best overall lung ...

A hybrid CNN–transformer model produces 128-dimensional medical image embeddings that stay robust under noise, blur, cropping, and occlusion across four ...