Recalculation of Lost Information in Neuron with Quadratic Spline Interpolation

Authors : Vedat Burak YÜCEDAĞ, İlker DALKIRAN

Pages : 132-137

Doi:10.31590/ejosat.1166055

View : 19 | Download : 15

Publication Date : 2022-09-30

Article Type : Research Paper

Abstract :The main function of neurons in a living creature is to transmit information. Neurons carry out information transmission without loss despite environmental and internal noise sources. However, sometimes there may be losses in the transmission of information. This results in diseases such as Alzheimer\`s, MS, and Epilepsy. In this study, the information lost in neurons is recalculated with the Quadratic Spline Interpolation method. In cases where it is difficult or impossible to calculate a function, the process of calculating the corresponding value of an unmeasured variable is called interpolation. In this study, first of all, three sample neuron behaviours are created with the Fitzhugh-Nagumo model, and the action potential and recovery parameter variables are obtained. Then, some data in the variables are deleted, resulting in unhealthy neuron behaviour. Then, these deleted data are recalculated using the Quadratic Spline Interpolation method. Various error values are obtained by comparing the actual and calculated data. The data lost in the action potential-recovery variable are detected with a very low error rate of 0.2630-0.0524%, 0.2885-0.0165% and 0.2543-0.0781% for the three sample neuron behaviours, respectively. With this study, it has been demonstrated that information lost or incorrectly coded in neurons for any reason can be corrected. It is also understood that this study can be used to prevent losses in real-time measurement results from biological neurons and to recalculate erroneous values.
Keywords : İkinci Dereceden İnterpolasyon, Kübik İnterpolasyon, FitzHugh Nagumo, Hodgkin Huxley, Nöron Model