Related Experiment Videos
BP neural network classification for bleeding detection in wireless capsule endoscopy
1School of Electronics, Information and Electrical Engineering, 820 Institute, Shanghai JiaoTong University, Shanghai, PR China. Guobpan@gmail.com
Insights
This study introduces an AI-powered method for detecting digestive tract bleeding in wireless capsule endoscopy images. The system accurately identifies bleeding regions, improving diagnostic efficiency and patient care.
Area of Science:
- Medical Imaging
- Artificial Intelligence
- Gastroenterology
Background:
- Digestive tract bleeding is a common and serious condition.
- Wireless capsule endoscopy (WCE) enables noninvasive visualization of the entire GI tract.
- Manual analysis of WCE images is time-consuming, hindering widespread clinical adoption.
Purpose of the Study:
- To develop an automated computer-aided detection (CAD) system for identifying bleeding in WCE images.
- To enhance the efficiency and accuracy of diagnosing gastrointestinal bleeding.
Main Methods:
- Extraction of color texture features from WCE images in RGB and HSI color spaces.
- Development of a neural network model utilizing these features for bleeding region recognition.
- Experimental validation of the proposed algorithm's performance.
Main Results:
- The developed algorithm successfully recognizes and delineates bleeding regions in WCE images.
- Achieved a sensitivity of 93% and a specificity of 96% for bleeding detection.
- Demonstrated the potential for accurate and automated bleeding identification.
Conclusions:
- The proposed AI-based method offers an effective solution for automatic bleeding detection in WCE.
- This technique can significantly reduce the manual workload associated with WCE image analysis.
- The high accuracy suggests potential for improved diagnosis and management of GI bleeding.
Abstract:
Bleeding in the digestive tract is one of the most common gastrointestinal tract (GI) diseases, as well as the complication of some fatal diseases. Wireless capsule endoscopy (WCE) allows physicians to noninvasively examine the entire GI tract. However it is very laborious and time-consuming to inspect large numbers of WCE images, which limits the wider application of WCE. It is therefore important to develop an automatic and intelligent computer-aided bleeding detection technique. In this paper, a new method aimed at bleeding detection in WCE images is proposed. Colour texture features distinguishing the bleeding regions from non-bleeding regions are extracted in RGB and HSI colour spaces; then a neural network using the colour texture features as the feature vector inputs is designed to recognize the bleeding regions. The experiments demonstrate that the bleeding regions can be correctly recognized and clearly marked out. The sensitivity of the algorithm is 93% and the specificity is 96%.
