Analysis Study Of Brain Electrical Impedance Tomography

Traumatic brain injury is a disease with rapid onset and high mortality.Clinically,biomedical imaging techniques such as CT and MRI are commonly used to identify craniocerebral injury.However,there are a few shortcomings such as radiation,large equipment,high cost,and not convenient of these traditional techniques.Electrical impedance tomography(EIT)is a new medical imaging technology that uses the different frequency responses of different biological tissues to measure the surface potential by applying a certain excitation to reverse the internal potential distribution.It is a non-invasive and rapid detection method,which has a good application prospect in the rapid detection of brain injury.During the process of brain data acquisition,the electrode lead of the electrode interface of the existing hardware platform was too long,which would cause the distributed capacitance of the electrode leads to a shunting effect on the excitation source.By adding a shielding drive circuit,the shunt effect of the electrode leads would be reduced,which could further improve the stability of the excitation source of the hardware system.During the process of detecting the connection between the electrodes of the hardware platform and the brain,it was very easy to cause data measurement errors due to detached electrodes.Meanwhile,it would cause the established model to not correspond well with the real boundary,resulting in artifacts in the imaging process,and even imaging failure when the measurement boundary model becomes inaccurate.In order to solve the above problems,this paper designed a hemispherical boundary contour detection device,which could fix the electrodes on the skin of the living body,and the boundary contour could be calculated by establishing a coordinate system.The experimental results showed that the designed device could fix the electrode plate better,and the error value of boundary measurement is small.Based on the circular field measured by the device,a division is performed to solve the conductivity distribution of the positive problem.In view of the difficulty in obtaining the data of the previous frame of brain injury,resulting in the inapplicability of differential imaging,the Newton-Raphson algorithm is used for image reconstruction.The NewtonRaphson algorithm can effectively obtain the imaging of a single frame of data.During the iterative process of the Newton-Raphson algorithm,it was found that the Hessian matrix has a high ill-conditioned,which makes the solution process of the inverse problem unstable.Therefore,a regularization term is added to constrain the matrix.The improvement of the objective function based on the original dual interior point method based on L1 norm and Tikhonov,NOSER and Laplace regularization based on L2 norm is studied to make the solution unique.which made the solution process of the inverse problem unstable.During single-target and dual-target simulation experiments,this paper established a brain simulation model with a three-layer structure of scalp,skull,and brain parenchyma,and then the L1 norm and the L2 norm were added to the static Newton-Raphson addition to the regularization term.The experimental results show that the algorithm reconstruction has a good effect,the image correlation of single-target imaging is 0.6558～0.6990,and the image correlation of dual-target imaging is 0.5662～0.6072.Finally,combined with the hardware system and designed structure,the human experiment data measurement is carried out.