طراحي شبكه عصبي مصنوعي ف بعنوان مدلي از شاخ خلفي ماد، خاكستري نخاع

Designing an Artificial Neural Network as a Model for Dorsal Horn of the Spinal Cord Grey Matter

Introduction: The purpose of this research was to elucidate the function of the dorsal horn of the spinal cord grey matter using an artificial neural network (ANN). Dorsal horn of the spinal cord grey matter, which is the site of termination of thin myelinated and unmyelinated afferent fibers, is influenced by thick myelinated fibers. The sophisticated structure of different layers of this part of the grey matter is one of the main causes of increased complexity in pain signal transmission and control. The dorsal horn of spinal cord grey matter modeling can be a potential method to elucidate mechanisms in this regard. In this study, we have evaluated the interaction of different peripheral inputs and the effects of CNS centers on the dorsal horn by designing an ANN. , Materials and Methods: A computer program was written in "C++" to design an ANN. This program produces a three layer ANN, which was trained by back propagation algorithm and the number of inputs, outputs and neurons can be determined by the user." MATLAB" software was used to evaluate the results and to plot the graphs. Results: The designed network had 5 neurons in each of the input and hidden layers. After training some specific features, the ANN was tested and the results confirmed complete learning. Then, we tested the ANN against some unlearned inputs and outputs. The results were consistent with acceptable learning. In other words, our ANN can predict the output against unlearned inputs. Conclusion: Compared with the previous models, our model has many advantages. The ANN method is a modern and simple modeling route with optional network structure, and has more input-output patterns. Our model can interpret rhythmic pain, the effects of stimulating thick neural fibers (touch) on the pain, and some other effects.