Related Experiment Video
Updated: Oct 11, 2025
08:05
Lung CT Segmentation to Identify Consolidations and Ground Glass Areas for Quantitative Assesment of SARS-CoV Pneumonia
Published on: December 19, 2020
14.4K
The Progress of Medical Image Semantic Segmentation Methods for Application in COVID-19 Detection
Amin Valizadeh1, Morteza Shariatee2
1Department of Mechanical Engineering, Ferdowsi University of Mashhad, Mashhad, Iran.
Computational Intelligence and Neuroscience
|November 29, 2021
Summary
This study reviews traditional and deep neural network methods for image semantic segmentation. It explores current deep learning advances for improved segmentation accuracy in medical imaging and computer vision.
Area of Science:
- Computer Vision
- Medical Imaging Analysis
- Artificial Intelligence
Background:
- Image semantic segmentation is crucial in medical imaging, computer vision, and intelligent transportation.
- Traditional methods and deep learning approaches are both utilized for image segmentation.
Purpose of the Study:
- To review traditional image segmentation methods.
- To explore current deep neural network-based semantic segmentation techniques.
- To provide a conclusion on the advancements in deep learning for semantic segmentation.
Main Methods:
- Review of traditional image segmentation methods and datasets.
- Exploration of deep neural network architectures including all-convolution networks, FCN with CRF, and pyramid methods.
- Analysis of various training strategies: supervised, semi-regulatory, and non-regulatory methods.
Main Results:
- Traditional methods and datasets for segmentation are reviewed.
- A comprehensive overview of deep neural network methods is presented, covering network structures, sampling, and feature integration.
- Various supervised, semi-regulatory, and non-regulatory approaches are discussed.
Conclusions:
- Deep neural network concepts offer significant advancements in semantic segmentation.
- Developed advances show promise for enhanced image analysis in various fields.
- Further research into deep learning is essential for improving segmentation.
