https://doi.org/10.1051/epjap/2014140130
Experimental study on light induced influence model to mice using support vector machine
College of Science, Nanjing University of Aeronautics and Astronautics, Nanjing
210016, P.R. China
a e-mail: nuaazhzhm@126.com
Received:
1
April
2014
Revised:
18
June
2014
Accepted:
19
June
2014
Published online:
12
August
2014
Previous researchers have made studies on different influences created by light irradiation to animals, including retinal damage, changes of inner index and so on. However, the model of light induced damage to animals using physiological indicators as features in machine learning method is never founded. This study was designed to evaluate the changes in micro vascular diameter, the serum absorption spectrum and the blood flow influenced by light irradiation of different wavelengths, powers and exposure time with support vector machine (SVM). The micro images of the mice auricle were recorded and the vessel diameters were calculated by computer program. The serum absorption spectrums were analyzed. The result shows that training sample rate 20% and 50% have almost the same correct recognition rate. Better performance and accuracy was achieved by third-order polynomial kernel SVM quadratic optimization method and it worked suitably for predicting the light induced damage to organisms.
© EDP Sciences, 2014
