Assessment of chronic kidney disease using skin texture as a key parameter: for South Indian population

Healthcare Technology Letters

article

Madhanlal Udhayarasu¹  Kalpana Ramakrishnan¹  Soundararajan Periasamy² 

[1] Department of Biomedical Engineering, Rajalakshmi Engineering College;Department of Nephrology, Sri Ramachandra University

关键词: diseases;    skin;    biological organs;    neural nets;    health care;    chronic kidney disease;    skin texture;    South Indian population;    glomerular filtration rate;    CKD epidemiology collaboration equations;    artificial neural network;   

DOI  :  10.1049/htl.2016.0098

学科分类：肠胃与肝脏病学

来源: Wiley

【 摘 要 】

Periodical monitoring of renal function, specifically for subjects with history of diabetic or hypertension would prevent them from entering into chronic kidney disease (CKD) condition. The recent increase in numbers may be due to food habits or lack of physical exercise, necessitates a rapid kidney function monitoring system. Presently, it is determined by evaluating glomerular filtration rate (GFR) that is mainly dependent on serum creatinine value and demographic parameters and ethnic value. Attempted here is to develop ethnic parameter based on skin texture for every individual. This value when used in GFR computation, the results are much agreeable with GFR obtained through standard modification of diet in renal disease and CKD epidemiology collaboration equations. Once correlation between CKD and skin texture is established, classification tool using artificial neural network is built to categorise CKD level based on demographic values and parameter obtained through skin texture (without using creatinine). This network when tested gives almost at par results with the network that is trained with demographic and creatinine values. The results of this Letter demonstrate the possibility of non-invasively determining kidney function and hence for making a device that would readily assess the kidney function even at home.

【 授权许可】

CC BY|CC BY-ND|CC BY-NC|CC BY-NC-ND