| Neuropsychiatric diseases such as depression and Alzheimer's disease have become important diseases that seriously affect human health,and the development of drugs for its treatment is progressing slowly.It has been proved that electrical,magnetic stimulation and light stimulation provide an effective means for the diagnosis and treatment of such diseases.Deep Transcranial Magnetic Stimulation(DTMS)is a new way to stimulate deep brain regions such as the hippocampus.It is an effective means to treat neuropsychiatric diseases caused by tissue diseases such as hippocampus.In order to obtain accurate positioning and appropriate dose,an accurate brain electric field distribution model is needed to achieve personalized electromagnetic stimulation.Therefore,this paper starts with the modeling of individual brain tissue,analyzes the correlation characteristics of DTMS coil based on personalized model,and proposes a method to obtain the magnetic field induced by magnetic stimulation in real time,and obtain accurate brain electric field distribution.First,based on MRI images,a real brain model was established using medical image software such as MIMICS.The grayscale threshold segmentation,morphological operation,region growth and manual editing were adopted.The reconstructed model contained five kinds of tissue structures: scalp,skull,cerebrospinal fluid,gray matter and white matter.Secondly,using the real brain model,the brain electromagnetic distribution of DTMS is obtained.The common DTMS coils were selected for simulation analysis.The characteristics of electromagnetic field distribution of DTMS coils under real brain model were analyzed by comparison with commonly used 8-shaped coils.The ratio of the deep electric field of the brain to the peak value of the electric field of the scalp was used as an evaluation index to characterize the depth characteristics of DTMS.Finally,the real-time electric field distribution prediction of magnetic stimulation is realized by deep learning.Based on the real brain model,the limitations of traditional research methods in clinical practice are analyzed,and a method for obtaining magnetic induction induced electric field in real time is proposed.With a deep neural network,the induced electric field can be predicted directly from the MRI image without establishing a physical model.The correlation between the method and the induced electric field obtained by the traditional research method is analyzed by corresponding indicators,and the reliability of the method is verified.In this paper,MRI image modeling is used to study brain tissue specificity,and the electromagnetic field distribution and depth characteristics of DTMS coils are analyzed.A method for real-time acquisition of magnetic stimulation induced electric field is proposed,which provides a theoretical basis for the application of DTMS in clinical practice. |
PDF Full Text Request