Hematocrit estimation using online sequential extreme learning machine

Hieu Trung Huynh, Quan Dac Ho


Hematocrit (HCT) is the volume percentage of red blood cells in the whole blood. This is the most highly influencing factor in glucose measurement using handheld devices. In this paper, we present a new approach to estimate hematocrit from the transduced current curve which is produced by chemical reaction on electrochemical biosensors used in glucose measurement. Our method utilizes the single-hidden layer feedforward neural network trained by online sequential extreme learning machine. The experimental results are given to show  high level of accuracy of the proposed method.


Hematocrit, neural network, extreme learning machine; online sequential training, glucose measurement.

DOI: https://doi.org/10.15625/1813-9663/29/3/2750

Journal of Computer Science and Cybernetics ISSN: 1813-9663

Published by Vietnam Academy of Science and Technology