Modeling the optical behavior of platelets in animal blood plasma using the turbidity method
Paper ID : 1447-NICAME1402
Authors:
sajjad izadi *1, alireza mahdavian1, saeed minaee2
1Department of Biosystem Mechanical Engineering, Faculty of Agriculture, Tarbiat Modares University, Tehran, Iran
2Biosystems Engineering Department, Faculty of Agriculture, Tarbiat Modares University (TMU), Tehran, Iran
Abstract:
Today, animal blood is used as an important biological product in the production of some strategic drugs; Considering the importance of blood products in the treatment process as well as the importance of the blood components detection system in the separation equipment, in this research, modeling of the optical behavior of platelets in animal blood plasma was done using the turbidity measurement method. To model the optical behavior of platelets, a multi-spectral optical system was designed based on the ISO standard. The wavelengths used in this research were selected from the ultraviolet to infrared spectrum. The reception of light at certain angles was determined in the three main directions of scattering, backscattering, and nephelometry. Modeling of the research was done by linear regression method in Python language. To pre-process the data and reduce the dimensions of the built model, two methods of principal component analysis and univariate feature selection were used. In this research, using modeling with the help of the dimensionality reduction method with less than 20% of the data in both dimensionality reduction methods, 95% of the variance of the total data was obtained. Absorption angles and smaller angles of 90 degrees were the most important features of the research angles. The most important feature among all angles and wavelengths was 750nm at zero degrees. The model compiled in this research is used in the production of higher quality blood products, online separation of blood components, and determination of platelets in blood samples.
Keywords:
Scattering, optical spectrum, backscattering, dimensionality reduction, nephrometry.
Status : Paper Accepted (Oral Presentation)