Effects of Dipole Model for Magnetic Induction on Biomedical Devices

In this dissertation, the magnetic field is analyzed for the development of medical applications using the extended distributed multipole method (eDMP). Then, the control method contributes to the formation of a magnetic field that improves the performance of applications. The eDMP method is used to analyze the magnetic field of the MIT system and implement the image reconstruction system.

Introduction

Motivation

Research tasks and contribution

Voice coil navigation sensor
Magnetic induction tomography

A navigation sensor has been widely used for the medical devices, such as minimally invasive surgical robotics [ 3 , 4 ], flexible catheters, injection needles and endoscopy [ 5 ], due to low cost, non-contact and harmless characteristics. In addition, an expert is required to operate the devices as they can be harmful. In contrast, MIT has low power and cost, and is therefore portable and can be used during emergencies.

Fig. 1.1. Utilization of intubation tube

Thesis organization

Literature reviews

Magnetic field analysis

Background of electromagnetic field
Distributed multi-pole model
Magnetic field control

Due to the fast and intuitive calculation, the equivalent circuit model is applied to calculate the system input current or voltage [8], and an equivalent inertia model calculates specific mechanical parameters, such as torque [9]. The dipole model is applied to calculate the current of needle-shaped coils, as shown in Fig. The lumped parameter model was then used to calculate the required current of each coil to control the magnetic field.

Fig. 2.1. Analytical model of magnetic field.

Biomedical applications

Navigation sensor
Magnetic induction tomography

For the iron object with triangle shape, imaging results showed its shape with three corners in Fig. showed. a) System setup (b) True image (c) Image result Fig. Jinxi Xiang obtained imaging for low conductivity objects using gradiometers [27]. Image reconstruction algorithm is one of the important parameters that determine image quality, since the MIT system has a bad condition during inverse, so image algorithms have been investigated [ 42 , 43 ]. However, this method faces difficulty in describing the detailed characters of the system, such as positions and orientations of Txs and Rxs.

Magnetic field modeling and control

Overview
Extended distributed multi-pole method

Magnetic field modeling
Magnetic induction

Validation and application

Experimental validation
Illustrative application: TMS

Magnetic field control

Magnetic field control using eDMP
Experimental validation
Illustrative application: Capsule endoscope

Discussion

The parameters of the eDMP can be calculated by minimizing the error between the predetermined magnetic field and the field described by (3.3) and (3.4). The magnetic field in the experiments was measured three times and the standard deviation of the mean was calculated in Table 3.2. Minimizing both the sum of the errors in (3.15) and the maximum error in (3.16) allows the magnetic field to be monitored at a specific location of the ROI.

This can be minimized by imposing a gradient magnetic field in (3.17) to compensate the error due to the discontinuity of the desired magnetic field, and reduce fluctuation in the field. 3.13 (c) shows the positions to control a desired magnetic field in the ROI, and the EMs similarly on the perimeter of the ROI. Unlike the uniform field, the intensity of the magnetic field along its discontinuity causes large errors due to mathematical singularity.

Comparison of the magnetic field distributions gives good agreement, indicating that the magnetic field can be controlled. 3.16 (b), the position of the capsule is controlled and displaced along the x-axis by a uniform gradient magnetic field ∂Bx/∂x. The position, orientation and strength of the EM array are determined to satisfy the desired magnetic field.

The usefulness and accuracy of the eDMP method and the controllability of the magnetic field have been demonstrated.

Fig. 3.1. eDMP model of a cylinder coil.

Voice coil navigation sensor

Overview
Magnetic field analysis

Design concept
Magnetic field modeling using eDMP
System design

Model simulation

Modeling of an inclined angle
Design optimization

Experiments and results

Experimental setup
Results and discussion
Application

Discussion

The induced voltages in the sensing coils and the variance of the mutual inductance when the bending angle changes should be analyzed to improve the sensing performance. The induced voltage in (4.1) depends on various geometrical design parameters such as the length of the coils, L and Lb, the distance l between the coils and the diameter D of the coils, as shown in Fig. To characterize the sensing performance, the resolution S and the linearity γ of the mutual inductance in terms of the tube bending angle  are defined in (4.5) and (4.6).

The overall magnitude of the mutual inductance decreases as the bending angle increases, especially after the angle reaches 45 degrees. The design parameters of the sense coil are summarized in Table 4.2 and the model parameters for the excitation coil are shown in Table 4.3. To increase the performance of the navigation sensor, the resolution for the tilt angle should be maximized.

The design parameters of the system are the coil length a and b, the distance between the two coils and the slope angle β, as shown in Fig. When the bending of the tube is performed within the slope surface of the inner coil, it is expected that the voltage induced in the inner coil is dominant compared to the outer coil. The performance of the navigation sensor with the design parameters is demonstrated numerically and experimentally in the range of tilt angle.

A set of excitation coils and inclined detection coils were wound along the tube and the induced voltage on the sensor was used to measure the curvature of the bent tube.

Magnetic induction tomography

Overview

Modeling using eDMP

Magnetic field of MIT
Conductivity estimation of MIT
Effect of objects on magnetic field
Performance estimation

The magnetic vector potential A, the flux density B and the mutual inductance between two coils can be calculated using the dipole moments in (3.3-4) and (3.7). For a given Tx and Rx coil design, the orientation and strength mi can be calculated using (5.1) and (3.4). As in (5.1), the induced field Ei in the coil Ii becomes perpendicular to the primary field on the ith object, Bik, which can be assumed to be uniform within the loop region Si (=πri2) surrounded by the current, especially for small ri.

Since Tx is activated by time harmonic input VIN, the Z parameter of the MIT system can be expressed as (5.4- 5.4b). The circuit parameters of the objects modeled as a single-turn coil, Ri and Li, can be calculated using the coil design shown in Fig. The various fields can be formed by activating the Tx arrangement with respect to the object and Rxs.

The mutual inductance Ml,o in the third row of (5.4a) can be neglected since the effect of the magnetic field of the Rx coil on the object is small, as the Rx produces a relatively weak field compared to the Tx coil. For each Tx and Rx, the phase shift due to the access Yo can be expressed using the sensitivity matrix P. Then the secondary field and the total field on the ith object in the nth calculation, Bio,n and Bik ,n, are estimated as shown in (5.16) and (5.17).

Resolutions can be defined as the variance of the parameters per unit phase shift.

Fig. 5.2. Equivalent circuit model of the MIT system.

Validation

MIT system setup
Simulation
Effect of conductivity variation
Effect of position
Effect of object shape
Sensitivity analysis
Image reconstruction

The circular object is placed at (0, 0.01) m and the eDMP model of the circle is shown in figure. Phase shift and CR increase with the square of the radius increase. a) eDMP model for object size (b) Phase shift as position changes. But the results have the same resolution as that of the eDMP model shown in Fig.

In the experimental images, the position of the objects corresponds to that of the real image. Image resolution can be improved by increasing the object dipoles in the vicinity of the object. Finally, the results show that MIT using eDMP can be applied to reconstruct the images to detect the conductivity, position and shape of the object. a) circle eDMP model (b) square eDMP model.

Due to frequency limitation of the MIT transceiver system, it should be used on high conductivity materials. 5.22 (a-b) shows the carbon rod with diameter 2.5 cm, thickness 1 mm and conductivity 3e5 S/m and the phase shift result It is located at the center of the ROI. Considering the crack detection application, the effect of the crack size on the system is investigated.

The results showed good agreement as the conductivity, position and shape of the object were changed.

Fig. 5.6. Block diagram of the MIT system.

Multi-channel system

Phase domain MIT transceiver
Multi-channel MIT system
System validation and analysis
Application: Carbon crack detection

Application: Egg imaging

It has conductivity about 0.3 S/m while albumin at room temperature is 0.7 S/m, due to fatty acid and water. RTx is reduced to 5.5 cm to increase SNR and Rxs with azimuth angle -30 to 210 deg are used to obtain the effective data. The object dipoles are distributed with ri = 2 mm and Ni = 439 per plane. a) Geometry of egg (b) Picture of egg.

The experimental image represents the shape of the egg, but the yolk is not marked due to lack of resolution, while the simulation in the image presents a challenge to observe the yolk, as the albumen signal is more dominant than the albumen signal. yolk. 5.29 (c) is smaller than the geometry in Table 5.7 because the cross section of the egg varies along the z axis.

Discussion

Conclusion and open issue

Achievement and contribution

A new approach to analyze the magnetic field of MIT using magnetic dipoles and a lumped parameter model was presented. It has a problem of worse performance due to non-linearity, low SNR and ill-posed mode in inverse problem. The eDMP model and a lumped parameter were used to solve the forward and inverse problem with the MIT system.

It was demonstrated that the system can characterize the conductivity, position and shape of the target object.

Open issue

Kim, “Low-Q electrically small spherical magnetic dipole antennas,” IEEE Transactions on Antennas and Propagation, vol. Simi et al., “Magnetically activated stereoscopic vision system for single-site laparoendoscopic surgery,” IEEE/ASME Transactions on Mechatronics, vol. 34;Electromagnetic Tracking in Medicine - A Review of Technology, Validation and Applications", IEEE Transactions on Medical Imaging, vol.

Conductivity Imaging in Mutual Induction Tomography for Industrial Applications", IEEE Transactions on Instrumentation and Measurement, vol. Zhang, "Magneto-acousto-electrical Measurement Based Electrical Conductivity Reconstruction for Tissues", IEEE Transactions on Biomedical Engineering, vol. Soleimani, "Theoretical and Experimentele evaluatie van rotatie-magnetische inductietomografie", IEEE Transactions on Instrumentation and Measurement, vol.

Gencer, "Electrical conductivity imaging via contactless measurements: an experimental study," IEEE Transactions on Medical Imaging, vol. Scharfetter, "Optimum Receiver Array Design for Magnetic Induction Tomography," IEEE Transactions on Biomedical Engineering, vol. Son, "Voice Coil Navigation Sensor for Flexible Silicone Intubation," IEEE/ASME Transactions on Mechatronics, vol.

Son, “Active control of magnetic field using eDMP model for biomedical applications,” IEEE/ASME Transactions on Mechatronics, vol.