Related Experiment Video
Updated: Jul 13, 2026
10:25
Deep Learning-Based Segmentation of Cryo-Electron Tomograms
Published on: November 11, 2022
10.5K
ResNext based U-Net for segmenting sonomammogram
Summary
This study introduces a novel ResNext-based U-Net for segmenting breast ultrasound images, improving accuracy in detecting breast cancer. The enhanced model aids radiologists in early diagnosis by providing reliable automated sonomammogram segmentation.
Area of Science:
- Medical Imaging
- Artificial Intelligence in Radiology
- Biomedical Engineering
Background:
- Breast cancer detection via ultrasound imaging faces challenges from variable image quality and interpretation.
- Accurate segmentation of sonomammograms is crucial for reliable computer-aided diagnosis.
Purpose of the Study:
- To introduce and evaluate a novel ResNext-based U-Net architecture for enhanced sonomammogram segmentation.
- To improve the accuracy and reliability of automated breast ultrasound image analysis.
Main Methods:
- Development of a U-Net architecture incorporating ResNext blocks for improved feature extraction.
- Integration of residual connections within the U-Net framework to enhance gradient propagation.
- Performance evaluation using five-fold cross-validation, comparing against baseline U-Net and ResNet encoders.
Main Results:
- The ResNext-based U-Net demonstrated improved segmentation performance compared to baseline models.
- The enhanced architecture showed particular effectiveness in capturing finer details within sonomammograms.
- Increased specificity in segmentation was observed with the proposed model.
Conclusions:
- The novel ResNext-based U-Net architecture offers a promising advancement for automated sonomammogram segmentation.
- This enhanced model can serve as a valuable tool for radiologists in early breast cancer detection and diagnosis.
- The study highlights the potential of integrating advanced deep learning techniques for more reliable medical image analysis.
