Related Experiment Video
Updated: Mar 27, 2026
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
Survey on Neural Networks Used for Medical Image Processing
Zhenghao Shi1, Lifeng He2, Kenji Suzuki3
1School of Computer Science and Engineering, Xi'an University of Technology, Xi'an, 710048, China, zhshi@ieee.org; Department of Radiology, The University of Chicago, 5841 South Maryland Avenue, MC 2026, Chicago, IL 60637, USA, suzuki@uchicago.edu; School of Computer Science and Engineering, Nagoya Institute of Technology, 464-8555, Japan, tnaka@juno.ics.nitech.ac.jp, itoh@juno.ics.nitech.ac.jp.
This review explores neural networks for medical image processing, classifying them by task and image type. It details their strengths, weaknesses, and future research directions in this field.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Computer Vision
Background:
- Medical image processing is crucial for diagnosis and treatment.
- Neural networks offer advanced capabilities for analyzing complex medical data.
- A comprehensive review is needed to guide research and application.
Purpose of the Study:
- To review neural network applications in medical image processing.
- To classify neural networks based on processing goals and image types.
- To identify current and future applications, strengths, and weaknesses.
Main Methods:
- Systematic literature review of neural networks in medical imaging.
- Classification of neural networks by processing objectives.
- Categorization based on the nature of medical images analyzed.
Main Results:
- Identified key applications of neural networks in medical image analysis.
- Detailed the advantages and disadvantages of various neural network approaches.
- Highlighted current challenges and future research opportunities.
Conclusions:
- Neural networks are powerful tools for medical image processing with diverse applications.
- Understanding their strengths and limitations is key for effective implementation.
- Future research should focus on addressing current challenges and expanding applications.