Related Experiment Video
Updated: Jan 10, 2026
05:56
Objectification of Tongue Diagnosis in Traditional Medicine, Data Analysis, and Study Application
Published on: April 14, 2023
3.2K
An Integrated Architecture for Colorectal Polyp Segmentation: The µ-Net Framework with Explainable AI.
Mehedi Hasan Emon1, Proloy Kumar Mondal1, Md Ariful Islam Mozumder1,2
1Institute of Digital Anti-Aging Healthcare, Inje University, Gimhae-Si 50834, Republic of Korea.
Diagnostics (Basel, Switzerland)
|November 27, 2025
Summary
This study introduces µ-Net, an AI tool that accurately detects and classifies polyps in colonoscopy images, improving early colorectal cancer (CRC) detection. Explainable AI enhances trust in its reliable performance for CRC screening.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Oncology
Background:
- Colorectal cancer (CRC) is a leading cause of cancer death globally.
- Early detection via colonoscopy is crucial for reducing CRC mortality.
- Manual polyp detection in colonoscopy is error-prone and inefficient.
Purpose of the Study:
- To develop an automated, reliable deep learning method for polyp segmentation and classification in colonoscopy.
- To enhance the accuracy and efficiency of colorectal cancer screening.
- To improve early detection rates and patient outcomes through AI-assisted analysis.
Main Methods:
- A novel deep learning architecture, µ-Net, was developed for polyp segmentation.
- The Kvasir-SEG dataset was used for training and evaluation.
- Explainable AI (XAI) techniques (saliency maps, Grad-CAM) were integrated for model interpretability.
Main Results:
- µ-Net achieved a high Dice coefficient of 94.02%, surpassing existing segmentation models.
- XAI techniques provided visual explanations, increasing confidence in the model's predictions.
- The model demonstrated strong accuracy and potential for clinical application.
Conclusions:
- The µ-Net framework offers a significant advancement in automated polyp screening.
- It improves the precision and efficiency of colonoscopy image analysis.
- This AI tool supports clinical decision-making for early CRC detection and prevention.
