The field of medical image analysis is rapidly advancing, with a focus on developing innovative methods for image segmentation, denoising, and feature extraction. Recent studies have proposed novel architectures and techniques, such as adaptive scan scores, attention mechanisms, and anatomy-aware discriminators, to improve the accuracy and robustness of medical image analysis. These advancements have the potential to enhance clinical diagnosis, treatment planning, and patient outcomes. Noteworthy papers in this area include: DBIF-AUNet, which achieved state-of-the-art performance in pleural effusion semantic segmentation, and VasoMIM, which introduced a novel masked image modeling framework for vessel segmentation. These papers demonstrate the significant progress being made in medical image analysis and highlight the potential for future innovations to transform the field.
Advances in Medical Image Segmentation and Analysis
Sources
A Semantic Segmentation Algorithm for Pleural Effusion Based on DBIF-AUNet
XAG-Net: A Cross-Slice Attention and Skip Gating Network for 2.5D Femur MRI Segmentation
An Implemention of Two-Phase Image Segmentation using the Split Bregman Method
VesselRW: Weakly Supervised Subcutaneous Vessel Segmentation via Learned Random Walk Propagation
Edge Detection for Organ Boundaries via Top Down Refinement and SubPixel Upsampling
LWT-ARTERY-LABEL: A Lightweight Framework for Automated Coronary Artery Identification
ASM-UNet: Adaptive Scan Mamba Integrating Group Commonalities and Individual Variations for Fine-Grained Segmentation
A Registration-Based Star-Shape Segmentation Model and Fast Algorithms
Anatomy-Aware Low-Dose CT Denoising via Pretrained Vision Models and Semantic-Guided Contrastive Learning
Enhanced Liver Tumor Detection in CT Images Using 3D U-Net and Bat Algorithm for Hyperparameter Optimization
SharpXR: Structure-Aware Denoising for Pediatric Chest X-Rays
Think as Cardiac Sonographers: Marrying SAM with Left Ventricular Indicators Measurements According to Clinical Guidelines
Glo-DMU: A Deep Morphometry Framework of Ultrastructural Characterization in Glomerular Electron Microscopic Images
FIND-Net -- Fourier-Integrated Network with Dictionary Kernels for Metal Artifact Reduction
VasoMIM: Vascular Anatomy-Aware Masked Image Modeling for Vessel Segmentation