Genetic algorithm search for the worst-case MRI RF exposure for a multiconfiguration implantable fixation system modeled using artificial neural networks.

PURPOSE: This paper presents a method to search for the worst-case configuration leading to the highest RF exposure for a multiconfiguration implantable fixation system under MRI. METHODS: A two-step method combining an artificial neural network and a genetic algorithm is developed to achieve this purpose. In the first step, the level of RF exposure in terms of peak 1-g and/or 10-g averaged specific absorption rate (SAR1g/10g ), related to the multiconfiguration system, is predicted using an artificial neural network. A genetic algorithm is then used to search for the worst-case configuration of this multidimensional nonlinear problem within both the enumerated discrete sample space and generalized continuous sample space. As an example, a generic plate system with a total of 576 configurations is used for both 1.5T and 3T MRI systems. RESULTS: The presented method can effectively identify the worst-case configuration and accurately predict the SAR1g/10g with no more than 20% of the samples in the studied discrete sample space, and can even predict the worst case in the generalized continuous sample space. The worst-case prediction error in the generalized continuous sample space is less than 1.6% for SAR1g and less than 1.3% for SAR10g compared with the simulation results. CONCLUSION: The combination of an artificial neural network with genetic algorithm is a robust technique to determine the worst-case RF exposure level for a multiconfiguration system, and only needs a small amount of training data from the entire system.

Fast Prediction of RF-induced Heating for Sacral Neuromodulation System Exposed to Multi-Channel 2 RF Field at 3T MRI.

This paper presents a fast method to predict the radiofrequency (RF) induced heating for Sacral Neuromodulation System (SNM) under multi-channel 2 (MC-2) RF field of 3 Tesla (T) magnetic resonance imaging (MRI) system by using the artificial neural network (ANN). The raw computational model for the SNM was based on the transfer function approach. The MC-2 parallel transmission RF field at 3T MRI exposure was considered for 2 independent channels, which have an exposure space of -15 dB to 15 dB magnitude difference and -180 degrees to 170 degrees phase difference. A total number of 535,680 study cases that cover all possible shimming conditions and the corresponding temperature rises are collected from raw calculation data. The ANN was used as the surrogate model to predict the temperature rises against the incident electromagnetic field distributions. 40320 cases were used for training while the rest data sets were used for testing. The ANN can estimate the temperature rises for each human model in a small exposure sampling space. The testing performance of the ANN has a correlation coefficient higher than 0.99 and the mean absolute error was less than 0.12°C. It is demonstrated that the ANN can be used as an efficient tool for quick temperature rise estimation under MRI 3T shimming.