@article { author = {Shahbazi, Sevda and Namdar, Aysan and Mesbahi, Asghar}, title = {Comparison between artificial neural network and radiobiological modeling for prediction of thyroid gland complications of after radiotherapy}, journal = {Iranian Journal of Medical Physics}, volume = {15}, number = {Special Issue-12th. Iranian Congress of Medical Physics}, pages = {241-241}, year = {2018}, publisher = {Mashhad University of Medical Sciences}, issn = {2345-3672}, eissn = {2345-3672}, doi = {10.22038/ijmp.2018.12871}, abstract = {Introduction: Hypothyroidism is one of the frequent side effects of radiotherapy of head and neck cancers, breast cancer, and Hodgkin's lymphoma. It is recommended to estimate the normal tissue complication probability of thyroid gland using radiobiological modeling during treatment planning. Moreover, the use of artificial neural network is also proposed as a new method for this aim. The purpose of this study is to compare the results of the radiobiological modeling with the artificial neural network. Materials and Methods: Thyroid dose-volume histograms from a dataset of 44 consecutive patients treated with 3D- CRT for head-and-neck cancers and breast cancer were extracted. Dose-volumetric and radiobiological parameters of thyroid gland including the minimum dose, the maximum dose, the mean dose, TD50, EUD, m, n, and NTCP values calculated by LKB model were used as the input variables of the artificial neural network (ANN) method. The results of each method were then compared with each other. Evaluation of the two methods was estimated using the root- mean-square error (RMSE) and the coefficient of determination (R2). Results: The values for simulation results were RMSE = 0.008 and R2 = 0.99. Using error estimation criteria, results obtained by the ANN method showed the potential prediction of this method for complications of thyroid gland as same as the radiobiological modeling method. Conclusion: Artificial neural network (ANN) can be used in modeling and patients could hopefully benefit from its individualized estimation of normal tissue complications in the clinic.}, keywords = {Radiobiological modeling,Artificial neural network,Prediction,Thyroid gland,NTCP,Radiotherapy}, url = {https://ijmp.mums.ac.ir/article_12871.html}, eprint = {} }