Kumar R.,Sathyabama University |
Ranganathan H.,Rajiv Gandhi College Engineering
Journal of Theoretical and Applied Information Technology | Year: 2014
Hemoglobin (Hb) is an imperative constituent of red blood cells (RBC). Assessment of individual physiological status is accomplished by measuring Hb concentration in human blood. The limitations associated with invasive methods are distress during sample collection, wide analysis time and disability to provide real time monitoring. These limitations were addressed by a gamut of non-invasive schemes. The decision scheme for different blood samples associated with non-invasive schemes exhibits intrinsic intricacies. This paper proposes a classification scheme for diverse blood samples by criterion based identification and association to an existing data set. The scheme was implemented and tested on three samples with different testing conditions. Accuracy was identified as the performance metric for this scheme and the results shows that the proposed schemes displays accuracy ranging from 99% to 72% for a range of 0 dB to 10dB Signal to Noise Ratio (SNR). © 2005 - 2014 JATIT & LLS. All rights reserved.