Creatinine prediction from body composition a neural network approach
Creatinine prediction from body composition a neural network approach
Files (excerpt)
Published date
2011
Resource type
Publisher
ISBN
ISSN
DOI
Call no.
Other identifier(s)
Edition
Copyrighted date
Language
eng
File type
application/pdf
Extent
8 pages
Other title(s)
Advisor
Other Contributor(s)
Citation
International Journal of Innovative Management, Information & Production
Volume 2, Number 1, March 2011
Degree name
Degree level
Degree discipline
Degree department
Degree grantor
Abstract
Creatinine, a naturally-produced chemical compound in blood, has been commonly used as a reliable indicator of kidney function. Creatinine level is typically obtained from blood-test. In this paper, a technique for predicting the criticality of creatinine level in blood is presented. The proposed technique takes only body size and mass parameters obtained from advanced weighing scale and body scanner, allowing the prediction to be done more casually. The technique applies a multi-layered feed-forward neural network for developing the prediction model. The achieved overall prediction accuracy is in the vicinity of 88% where the average false negative rate and the average false positive rate are 22.15% and 8.26%, respectively.