TY - GEN
T1 - Stage classification in chronic kidney disease by ultrasound image
AU - Hsieh, Jun Wei
AU - Lee, C. Hung
AU - Chen, Y. Chih
AU - Lee, W. Shan
AU - Chiang, H. Fen
PY - 2014/11/19
Y1 - 2014/11/19
N2 - Ultrasound imaging can provide radiation-free, non-invasive, low cost, and convenient for disease detection. However, speckle effect makes it noisy and thus reduces its overall diagnostic abilities in disease analysis. This paper develops a real time system to analyze chronic kidney disease (CKD) using only Ultrasound images. As we know, this is the first work to analyze CKD stages of patients directly from ultrasound images without using any blood examination such as Creatinine index. To build the scoring index, this paper uses Nakagami distribution and Local Binary Pattern (LBP) to model the scattering properties of CKD patients' ultrasound images. In addition, we find the age distribution is also important for CKD stage analysis. After integration, a codebook concept is adopted to extract important visual codes to describe various texture and scattering characteristics of each CKD stage. Then, an ensemble scheme is proposed for CKD stage prediction and classification by separating input ultrasound images to several grids and then integrating different classifiers trained on these grids to build a strong CKD stage classifier via SVM. Experimental results demonstrate the sensitivity and specificity of this system up to 93.82% and 83.34%, respectively.
AB - Ultrasound imaging can provide radiation-free, non-invasive, low cost, and convenient for disease detection. However, speckle effect makes it noisy and thus reduces its overall diagnostic abilities in disease analysis. This paper develops a real time system to analyze chronic kidney disease (CKD) using only Ultrasound images. As we know, this is the first work to analyze CKD stages of patients directly from ultrasound images without using any blood examination such as Creatinine index. To build the scoring index, this paper uses Nakagami distribution and Local Binary Pattern (LBP) to model the scattering properties of CKD patients' ultrasound images. In addition, we find the age distribution is also important for CKD stage analysis. After integration, a codebook concept is adopted to extract important visual codes to describe various texture and scattering characteristics of each CKD stage. Then, an ensemble scheme is proposed for CKD stage prediction and classification by separating input ultrasound images to several grids and then integrating different classifiers trained on these grids to build a strong CKD stage classifier via SVM. Experimental results demonstrate the sensitivity and specificity of this system up to 93.82% and 83.34%, respectively.
KW - Chronic kidney disease
KW - Local Binary Pattern
KW - Nakagami distribution
KW - Support vector machine
UR - http://www.scopus.com/inward/record.url?scp=84985898416&partnerID=8YFLogxK
U2 - 10.1145/2683405.2683457
DO - 10.1145/2683405.2683457
M3 - 会议稿件
AN - SCOPUS:84985898416
T3 - ACM International Conference Proceeding Series
SP - 271
EP - 276
BT - Proceedings of IVCNZ 2014
PB - Association for Computing Machinery
T2 - 29th International Conference on Image and Vision Computing New Zealand, IVCNZ 2014
Y2 - 19 November 2014 through 21 November 2014
ER -