Related Experiment Video
Updated: Jun 8, 2025
04:48
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
Published on: November 30, 2022
2.7K
Deep learning in image segmentation for cancer
Robba Rai1,2,3
1South Western Sydney Clinical School, University of New South Wales, Liverpool, New South Wales, Australia.
Journal of Medical Radiation Sciences
|November 6, 2024
Summary
Deep learning (DL) significantly enhances cancer imaging by automating image segmentation for faster, more accurate analyses. Further research is needed to overcome challenges like variable image quality across systems.
Area of Science:
- Medical Imaging
- Artificial Intelligence in Oncology
- Radiology
Background:
- Deep learning (DL) models, particularly U-Net and convolutional neural networks (CNNs), are increasingly applied in medical imaging for automated segmentation tasks.
- Accurate segmentation is crucial for quantitative analysis and treatment planning in oncology.
Discussion:
- This article reviews DL applications in cancer imaging, focusing on automatic segmentation for improved diagnostic accuracy and efficiency.
- Two studies showcase DL's success in body composition analysis using CT scans and rectal tumor segmentation via MRI.
Key Insights:
- DL-based architectures like U-Net and CNNs demonstrate potential to enhance the speed and precision of image segmentation in cancer diagnostics.
- Successful applications include body composition analysis in CT and rectal tumor delineation in MRI, improving quantitative assessments.
Outlook:
- Further research is essential to standardize DL algorithms and mitigate issues related to image quality variations across diverse imaging modalities and manufacturers.
- Addressing these challenges will be key to the widespread clinical adoption of DL for cancer imaging analysis.
