TY - GEN
T1 - Texture analysis of ultrasound images of chronic kidney disease
AU - Iqbal, Fadil
AU - Pallewatte, Aruna S.
AU - Wansapura, Janaka P.
N1 - Publisher Copyright:
© 2017 IEEE.
PY - 2017/7/1
Y1 - 2017/7/1
N2 - Chronic Kidney Disease of unknown aetiology (CKDu) is a prevalent disease in the North Central Province of Sri Lanka. Towards the latter stages of the disease, kidney function fails by 80%. During the initial stages of CKDu, interstitial fibrosis is formed and grows as the disease progresses. The cause of the disease remains elusive and early detection is vital to arrest the progressive decline of kidney function. The objective of this study is to construct a computer program to perform texture analysis on ultrasound kidney images and extract various features that can be used to distinguish between normal and diseased kidney patients. The computer program was developed using MATLAB and a user interface was created to perform mathematical operations such as: Fourier analysis to extract Root Mean Square and First Moment values and Grey Level Co-occurrence Matrix (GLCM) to extract Homogeneity and Sum Average values. A sample of ultrasound images were taken from 32 patients. Region of interest (ROI) selection was performed on entire kidney, cortex region and white (renal medulla or renal sinus) region separately. Among these methods Root Mean Square values over the entire kidney (p=0.03) and cortex region (p=0.0049) gave significant results in distinguishing between normal and diseased kidneys.
AB - Chronic Kidney Disease of unknown aetiology (CKDu) is a prevalent disease in the North Central Province of Sri Lanka. Towards the latter stages of the disease, kidney function fails by 80%. During the initial stages of CKDu, interstitial fibrosis is formed and grows as the disease progresses. The cause of the disease remains elusive and early detection is vital to arrest the progressive decline of kidney function. The objective of this study is to construct a computer program to perform texture analysis on ultrasound kidney images and extract various features that can be used to distinguish between normal and diseased kidney patients. The computer program was developed using MATLAB and a user interface was created to perform mathematical operations such as: Fourier analysis to extract Root Mean Square and First Moment values and Grey Level Co-occurrence Matrix (GLCM) to extract Homogeneity and Sum Average values. A sample of ultrasound images were taken from 32 patients. Region of interest (ROI) selection was performed on entire kidney, cortex region and white (renal medulla or renal sinus) region separately. Among these methods Root Mean Square values over the entire kidney (p=0.03) and cortex region (p=0.0049) gave significant results in distinguishing between normal and diseased kidneys.
KW - Chronic kidney disease
KW - Co-occurrence matrix
KW - Fourier analysis
KW - Texture analysis
KW - Unknown aetiology
UR - http://www.scopus.com/inward/record.url?scp=85049420140&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85049420140&partnerID=8YFLogxK
U2 - 10.1109/ICTER.2017.8257787
DO - 10.1109/ICTER.2017.8257787
M3 - Conference contribution
AN - SCOPUS:85049420140
T3 - 17th International Conference on Advances in ICT for Emerging Regions, ICTer 2017 - Proceedings
SP - 299
EP - 303
BT - 17th International Conference on Advances in ICT for Emerging Regions, ICTer 2017 - Proceedings
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 17th International Conference on Advances in ICT for Emerging Regions, ICTer 2017
Y2 - 7 September 2017 through 8 September 2017
ER -