Using SAVE for Esophageal Cancer Detection
Author Information
Author(s): Chen Yen-Chun, Karmakar Riya, Mukundan Arvind, Huang Chien-Wei, Weng Wei-Chun, Wang Hsiang-Chen
Primary Institution: Dalin Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation
Hypothesis
Can the SAVE algorithm improve the detection of esophageal cancer using WLI images?
Conclusion
The SAVE algorithm produces images with better precision and F1-scores compared to traditional WLI images for detecting esophageal cancer.
Supporting Evidence
- The SAVE algorithm was developed to simulate NBI from WLI images.
- Results showed that SAVE-NBI images had superior accuracy compared to WLI images.
- The SSD model outperformed YOLOv5 and YOLOv8 in detecting cancerous images.
- Using SAVE technology improved the visibility of cancerous regions in images.
Takeaway
This study created a new way to help doctors see cancer better by using special computer techniques on regular images.
Methodology
The study used a dataset of WLI and NBI images, applying the SAVE algorithm to convert WLI images into simulated NBI images for analysis.
Limitations
The study's findings may not be generalizable to all healthcare settings due to the specialized nature of NBI equipment.
Digital Object Identifier (DOI)
Want to read the original?
Access the complete publication on the publisher's website