MRI-Based Amyloid Beta Estimation Using Trained Models
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for detecting early signs of Alzheimer's disease and other amyloid β-related diseases rely on PET images, which require radioactive substances and may not be suitable for patients with certain conditions, posing risks and limitations.
Innovation Solution
A system utilizing MRI images and a trained model to generate correlation images between magnetic susceptibility and amyloid β, allowing for the estimation of disease signs without PET images, using quantitative susceptibility mapping and machine learning techniques.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If PET images are used to detect early signs of Alzheimer's disease, then measurement precision is improved, but object-affected harmful factors increase due to radiation exposure from radioactive substances
Solution Approach 1:
The patent replaces the PET imaging system (which uses radioactive substances) with an MRI-based quantitative susceptibility mapping system. This substitution eliminates radiation exposure while maintaining the ability to detect amyloid β deposition through magnetic susceptibility measurements of iron in the brain.
Solution Approach 2:
The patent uses iron deposition as an intermediary marker to indirectly detect amyloid β. Instead of directly imaging amyloid β with radioactive tracers, the system measures iron accumulation (which correlates with amyloid β deposition) using MRI's quantitative susceptibility mapping, thereby avoiding direct radioactive exposure.
2Measurement precision
If PET images are used for disease detection, then measurement precision is improved, but adaptability worsens due to contraindications for patients with certain diseases
Solution Approach 1:
The patent replaces the PET imaging system with an MRI-based system. MRI is inherently safer and can be used by patients with kidney diseases and other conditions where PET with radioactive substances is contraindicated, thereby improving adaptability while maintaining diagnostic capability.
3Object-affected harmful factors
If MRI images and trained models are used to estimate disease signs, then object-affected harmful factors are reduced, but manufacturing precision worsens due to reliance on trained models
Solution Approach 1:
The patent performs preliminary training of machine learning models using datasets with known outcomes. This preliminary action creates a trained model that can accurately predict amyloid β deposition from MRI images, ensuring prediction accuracy is established before actual diagnostic use.
Solution Approach 2:
The patent creates a computational model that copies the relationship between MRI signal characteristics and amyloid β deposition patterns. This virtual model allows accurate prediction without physical radioactive exposure, transferring the diagnostic function from physical PET imaging to computational analysis.
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
Enables the estimation of early signs of amyloid β-related diseases, such as dementia, without the need for PET images, reducing radiation exposure and expanding suitability for patients with certain conditions, while providing accurate predictions of disease progression.
Implementation Method 1
a magnetic susceptibility capable of being specified on the basis of the MRI image
Data Source
AI summary
To provide a recording medium capable of estimating the early signs of diseases relating to amyloid β without using a PET image, an information processing device, an information processing method, a trained model generation method, and a correlation image output device.The recording medium recording the computer program causes a computer to execute processes of: acquiring an MRI image of a subject; and inputting the acquired MRI image to a trained model that outputs a correlation image representing a correlation between a magnetic susceptibility capable of being specified on the basis of the MRI image and amyloid β in a case where the MRI image is input, and outputting the correlation image representing the correlation between the magnetic susceptibility of the subject and amyloid β.