TY - GEN
T1 - Bleeding Region Segmentation in Wireless Capsule Endoscopy Images by K-Mean Clustering Technique
AU - Seebutda, Areeya
AU - Sakuncharoenchaiya, Sirilak
AU - Numpacharoen, Kawee
AU - Wiwatwattana, Nuwee
AU - Charoen, Amber
AU - Charoenpong, Theekapun
N1 - Publisher Copyright:
© 2023 IEEE.
PY - 2023
Y1 - 2023
N2 - Wireless capsule endoscopy (WCE) is used to record internal images of the gastrointestinal tract. A common symptom such as gastrointestinal bleeding can be diagnosed by images. In this paper, we proposed a method for bleeding region in gastrointestinal segmentation by the K-Mean Clustering technique. The images were captured by wireless capsule endoscopy (WCE). This method consists of three steps: preprocessing, color clustering, and bleeding region segmentation. Firstly, input data in RGB color space is converted to L*a*b∗ color space. Color intensity has two cluster which is bleeding region, and background. The K-Mean technique is used to group the data. Finally, bleeding region is defined by intensity in red layer. In experimental result, 48 images from KID Atlas dataset are used. The accuracy rate is 84.26%, DICE rate is 67.71%, Jaccard Index (JI) is 60.43%, sensitivity rate is 69.84% and the precision rate is 65.70%. The results is satisfactory for future improvement.
AB - Wireless capsule endoscopy (WCE) is used to record internal images of the gastrointestinal tract. A common symptom such as gastrointestinal bleeding can be diagnosed by images. In this paper, we proposed a method for bleeding region in gastrointestinal segmentation by the K-Mean Clustering technique. The images were captured by wireless capsule endoscopy (WCE). This method consists of three steps: preprocessing, color clustering, and bleeding region segmentation. Firstly, input data in RGB color space is converted to L*a*b∗ color space. Color intensity has two cluster which is bleeding region, and background. The K-Mean technique is used to group the data. Finally, bleeding region is defined by intensity in red layer. In experimental result, 48 images from KID Atlas dataset are used. The accuracy rate is 84.26%, DICE rate is 67.71%, Jaccard Index (JI) is 60.43%, sensitivity rate is 69.84% and the precision rate is 65.70%. The results is satisfactory for future improvement.
KW - Clustering
KW - Gastrointestinal bleeding
KW - K-mean
KW - Segmentation
KW - Wireless capsule endoscopy
UR - http://www.scopus.com/inward/record.url?scp=85149652035&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85149652035&partnerID=8YFLogxK
U2 - 10.1109/ICA-SYMP56348.2023.10044741
DO - 10.1109/ICA-SYMP56348.2023.10044741
M3 - Conference contribution
AN - SCOPUS:85149652035
T3 - 2023 3rd International Symposium on Instrumentation, Control, Artificial Intelligence, and Robotics, ICA-SYMP 2023
SP - 69
EP - 72
BT - 2023 3rd International Symposium on Instrumentation, Control, Artificial Intelligence, and Robotics, ICA-SYMP 2023
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 3rd International Symposium on Instrumentation, Control, Artificial Intelligence, and Robotics, ICA-SYMP 2023
Y2 - 18 January 2023 through 20 January 2023
ER -