Estimating Tissue Conductivity via MRI and Surface Potential
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for estimating conductivity distribution in body tissues, such as Electrical Property Tomography (EPT) and Electrical Impedance Tomography (EIT), face limitations in accuracy and invasiveness, particularly at low frequencies and in characterizing internal tissue conductivity without direct current injection.
Innovation Solution
A computer-implemented method combining magnetic resonance imaging (MRI) and body surface electrical potential measurements to estimate conductivity distribution non-invasively, using a neural network to solve the inverse problem and account for tissue morphology, thereby improving accuracy and avoiding current-based artifacts.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If Electrical Impedance Tomography (EIT) techniques are used to image biological tissues according to their electrical properties, then the estimation of body electrical property map can be performed, but the inverse problem is ill-posed and requires complex Finite Element Methods (FEM) which increases computational complexity
Solution Approach 1:
The patent replaces the traditional FEM-based computational approach with a deep learning model that uses neural networks to solve the inverse problem. The model takes surface electrical potential measurements and MRI images as input and directly outputs the conductivity distribution, avoiding the need for complex iterative FEM calculations while maintaining estimation accuracy.
Solution Approach 2:
The patent creates a digital copy of the tissue conductivity distribution through training the neural network on simulated data. Once trained, the model can rapidly infer conductivity maps from new surface measurements without re-solving the complex inverse problem, effectively copying the computational burden during training to enable fast inference during application.
2Ease of operation
If MRI-based EPT techniques are used to image electrical properties, then non-invasive imaging is achieved, but the performance is highly dependent on stimulation frequency and tissues lack characterization at low frequencies
Solution Approach 1:
The patent merges surface electrical potential measurements (which provide direct electrical property information) with MRI images (which provide morphological information) to create a comprehensive conductivity distribution map. This combination allows the system to overcome the frequency limitations of EPT by using electrical measurements that are valid across different frequencies while maintaining non-invasive imaging capability.
Solution Approach 2:
The patent uses deep learning as an intermediary that bridges the gap between surface electrical measurements and internal tissue conductivity. The neural network model processes both MRI and surface potential data to infer the conductivity distribution, acting as a mediator that translates surface measurements into internal property maps without requiring direct tissue penetration or frequency-dependent stimulation.
3Measurement precision
If direct current injection is used to measure electrical properties, then accurate conductivity measurement is achieved, but the method becomes invasive and introduces current-based artifacts
Solution Approach 1:
The patent uses the MRI system's own electromagnetic fields to induce surface electrical potentials that can be measured non-invasively. Instead of requiring external current injection, the system leverages the existing MRI electromagnetic fields to generate the necessary electrical signals for conductivity measurement, making the process self-service and non-invasive while avoiding current-based artifacts.
Solution Approach 2:
The patent substitutes direct current injection with electromagnetic field-based potential measurement. Instead of physically injecting current through electrodes, the system uses MRI-generated electromagnetic fields to induce surface potentials that can be measured without skin penetration, replacing the invasive electrical injection mechanism with a non-invasive electromagnetic measurement approach.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
The method provides a reliable, non-invasive estimation of conductivity distribution across body tissues, enhancing accuracy by integrating morphological and electrical data, and reducing computational complexity compared to traditional finite element methods.
Implementation Method 1
obtaining a magnetic resonance image (MRI) of the portion of tissue
Implementation Method 2
obtaining a body surface electrical potential measurement associated with the MRI
Data Source
Figure 1
Figure 2
Figure 3
AI summary
It is provided a computer-implemented method for estimating a distribution of conductivity over a portion of a tissue of a body. The method comprises obtaining a magnetic resonance image (MRI) of the portion of tissue and a body surface electrical potential measurement associated with the MRI. The method also comprises estimating the distribution of conductivity on the portion of tissue based on the body surface electrical potential measurement and on the MRI. This forms an improved solution for estimating the distribution of conductivity over a portion of a tissue of a body.