MRI Habitat Analysis Identifies High-Risk Prostate Cancer
A two-center study developed a biparametric MRI habitat analysis framework that effectively identifies pathologically high-risk prostate cancer, showing superior discrimination ...
Tag: Medical Imaging
Feasibility Study Demonstrates AI-Assisted 3D Stomach Reconstruction From Routine Gastroscopy
A retrospective feasibility study found that a Dense-UNet model reconstructed 3D stomach models from routine gastroscopy videos of eight patients ...
BigpictureQC: multi-stain, multi-species artifact segmentation for whole-slide quality control
A new deep learning framework called BigpictureQC automates the detection of artifacts in digital pathology slides, identifying poor-quality images with ...
Thyroid nodule AI segmentation models falter across medical centers
A comparative study found all evaluated thyroid nodule ultrasound segmentation models declined substantially across centers, with TNet achieving the best ...
Lightweight AI Accurately Identifies Pediatric Lung Infiltrates on Chest Radiographs
A lightweight MobileNet V2 deep learning model detected pediatric pulmonary infiltrates with 90 percent accuracy and 95 percent sensitivity while ...
Improved 2.5D-UNet Enhances Pseudo-CT Generation for Head PET Attenuation Correction
A 2.5D-UNet architecture using consecutive PET slices outperforms standard 2D models in generating pseudo-CT scans for PET attenuation correction.
MRI Radiogenomic Score Captures Tumor Microenvironment Signatures in Nasopharyngeal Carcinoma
Researchers developed a contrast-enhanced MRI radiogenomic surrogate reflecting a composite hypoxia, angiogenesis, and epithelial–mesenchymal transition transcriptional state that associates with ...
Wavelet Network Enhances Small Vascular Segmentation in Medical Angiography
Researchers developed a wavelet-bilateral gated transfer network that improved automated segmentation of small-scale vascular structures on clinically acquired computed tomography ...
Degraded Chest Radiographs Reduce Diagnostic Sensitivity in Deep Learning Tuberculosis Detection Models
Deep learning models for tuberculosis screening suffer substantial diagnostic sensitivity losses and increased calibration error when evaluated on chest radiographs ...
CT Histogram Radiomics Accurately Distinguishes Middle Ear Cholesteatoma
A retrospective study demonstrates that histogram-based radiomic texture analysis of temporal bone HRCT scans accurately differentiates middle ear cholesteatoma from ...
Machine Learning Model Integrates Multimodal Data to Stratify Glioblastoma Survival Risk
A multimodal machine learning nomogram combining clinical, radiomics, and pathomics features improved overall survival risk stratification in patients with IDH-wildtype ...
Systematic review maps the state of AI deepfake detection across modalities and domains
A PRISMA-based systematic review of 142 primary studies categorises AI deepfake generation and detection methods and identifies dataset bias, poor ...
Benchmarking Deep Transfer Learning Models for COVID-19 Lung Ultrasound Detection
A systematic benchmarking study demonstrates that DenseNet-121 paired with the Adam optimizer achieves exceptional diagnostic accuracy in detecting COVID-19 from ...
Deep Learning Synthesizes Cross-Contrast MRI Scans Across Thirty Reviewed Medical Imaging Studies
A systematic review of thirty studies reveals that generative adversarial networks and diffusion models lead deep learning efforts to synthesize ...
Mammography may flag heart risk: breast calcifications tied to menopause and cholesterol
In a prospective case-control study, women with breast arterial calcifications had a younger menopausal age and more pregnancies, while calcification ...
AI learns to map children’s swallowing videos from sparse expert labels
Deep learning models trained on sparse expert annotations can segment anatomical structures in pediatric swallowing endoscopy videos from unseen patients, ...
Low Tube Voltage and High-Concentration Contrast Media Optimize One-Stop CT Angiography
A low tube voltage protocol combined with high-concentration contrast media reduces radiation dose during one-stop head-neck and coronary CT angiography ...
AI Learns When It Is Unsure: New Model Flags Borderline Hip Screenings Like Clinicians Do
Oxford researchers built an AI method that estimates uncertainty in landmark-based hip measurements and converts it into a screening confidence ...
Adaptive Decoder Learns Where to Fuse Features in Medical Image Segmentation
Researchers have developed DynamicFusionNet, a decoder that adaptively fuses features across layers, scales and views to improve medical image segmentation ...
Trained students can match radiologists in knee ultrasound ratings, study finds
A study found that groups of trained student raters, when aggregated, can approach expert radiologist accuracy in segmenting knee recess ...
POPULAR NEWS
-
Scoping Review Maps Continuous Glucose Monitoring Research in Adults With Prediabetes
-
Ethnographic Study Evaluates Forces Shaping Hip Fracture Care in Zimbabwe
-
Computational Study Shows Proteomes Can Preserve Folds Without Glutamine
-
Desi Chickpea Varieties Identified for Yield Stability Across Northwest Ethiopian Environments