Pattern Classification (2nd Edition)
Pattern Classification (2nd Edition)
Numerical analysis of factors which influent the biotic systems using the ferment activity
SENSIG'08 Proceedings of the 1st WSEAS international conference on Sensors and signals
Hi-index | 0.00 |
In measuring certain biological and chemical information from the human whole blood, several parameters are involved. The measurement parameters for the whole blood are varies from one sample to another, which makes measurement of a biological parameter less consistent through repeated measurements. In this situation, it is very hard to perform a total quality assurance for a bio-optical signal acquisition system due to variation over the sample measurements because many types of error can be caused at any step of the measurement procedure. In this paper, we propose a clustering-based measurement for quality assurance for a bio-optical signal acquisition system. We defined 'Discrimination Index' used as an index for quality verification of the system. We present an experimental result for a measurement system for a bio-optical glucose sensor. For three groups of human blood samples, the experiment showed that our proposed method produced distinct values for two different sample groups, while produced similar values for two similar sample groups. Our proposed method can be also used as a generalized method for reliability verification for a signal acquisition system that measures the samples with parameters that varies from one sample to another.