Related Experiment Video
Updated: Jun 14, 2025
04:48
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
2.7K
Insights and Considerations in Development and Performance Evaluation of Generative Adversarial Networks (GANs): What
Jeong Taek Yoon1, Kyung Mi Lee1, Jang-Hoon Oh1
1Department of Radiology, Kyung Hee University Hospital, Kyung Hee University College of Medicine, 23 Kyungheedae-ro, Dongdaemun-gu, Seoul 02447, Republic of Korea.
Diagnostics (Basel, Switzerland)
|August 29, 2024
Summary
Generative adversarial networks (GANs) offer solutions for deep learning in medical imaging, reducing the need for extensive labeled data. These advanced AI models enhance image augmentation, reconstruction, and anomaly detection for radiologists.
Area of Science:
- Artificial Intelligence
- Medical Imaging
- Deep Learning
Background:
- Deep learning in medical imaging faces challenges with large labeled datasets and manual segmentation.
- Generative adversarial networks (GANs) provide synthetic data generation for augmentation and streamline image processing.
- GANs enable unsupervised anomaly detection, reducing dependency on labeled data.
Purpose of the Study:
- To explore Generative Adversarial Networks (GANs) in medical imaging for radiologists.
- To provide a comprehensive understanding of GAN architectures, selection, training, and evaluation.
- To guide practical application and evaluation of GANs in brain imaging.
Main Methods:
- Investigated various GAN architectures (cGAN, CycleGAN, StyleGAN).
- Focused on data augmentation, image reconstruction, and segmentation.
- Utilized CycleGAN and pSp-combined StyleGAN for practical brain imaging examples.
Main Results:
- GANs improve efficiency in medical image augmentation, reconstruction, and segmentation.
- GANs facilitate unsupervised anomaly detection, reducing the need for labeled datasets.
- Demonstrated practical applications of GANs in brain imaging.
Conclusions:
- GANs offer significant advancements for medical imaging research and practice.
- This work equips radiologists with knowledge for effective GAN utilization.
- Encourages further research and application of GANs in the medical field.
