Breast MR Mammography Using Multi-Echo Deep-Learned Tissue Maps
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional MR imaging for breast cancer detection faces challenges in distinguishing between healthy tissue and cancerous abnormalities, leading to false positives and inability to identify tiny calcium deposits, limiting its effectiveness compared to conventional mammography.
Innovation Solution
A method involving RF excitation pulses, magnetic field gradients, and deep learning algorithms to generate and process echo signals along radial or spiral k-space trajectories, deriving high-resolution water, fat, B0, and T2* maps, and synthesizing a high-resolution mammogram with Hounsfield units for accurate tissue and calcification mapping.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Object-affected harmful factors
If conventional MR imaging is used for breast cancer detection, then non-ionizing radiation and soft tissue visualization are improved, but the ability to distinguish healthy tissue from cancerous abnormalities and identify micro calcifications deteriorates
Solution Approach 1:
The patent segments the MR imaging process into multiple echo signal acquisitions at different echo times, allowing separate optimization for different tissue characteristics. This enables simultaneous improvement of soft tissue visualization and micro calcification detection without compromising either aspect
Solution Approach 2:
The patent introduces a new temporal dimension by acquiring multiple echo signals at different echo times (TE1, TE2, TE3). This additional time dimension allows the system to differentiate between tissues based on their relaxation characteristics while maintaining the non-ionizing radiation advantage of MR imaging
2Reliability
If conventional MR imaging parameters are used, then overall image quality is maintained, but the resolution for detecting micro calcifications and tiny lesions deteriorates
Solution Approach 1:
The patent applies local quality enhancement by deriving separate high-resolution maps for different tissue types (water, fat, B0, T2*) and calculating Hounsfield units specifically for mammographic visualization. This allows micro calcifications to be detected with high precision while maintaining overall image quality through localized optimization
Solution Approach 2:
The patent changes the parameter representation by calculating Hounsfield units from the MR data, enabling the use of familiar mammographic contrast characteristics. This parameter transformation allows the system to achieve micro calcification detection resolution comparable to conventional mammography while maintaining the benefits of MR imaging
3Measurement precision
If multiple echo signals are acquired at different echo times along radial or spiral k-space trajectories, then micro calcification identification is improved, but the imaging time and data processing complexity increase
Solution Approach 1:
The patent uses periodic echo signal acquisitions at different echo times with radial or spiral k-space trajectories. This periodic sampling pattern allows efficient coverage of k-space while maintaining the ability to distinguish micro calcifications, balancing acquisition time with diagnostic precision
Solution Approach 2:
The patent performs preliminary k-space trajectory planning and echo time optimization before actual imaging. By pre-calculating the optimal radial or spiral trajectories and echo time sequence, the system minimizes acquisition time while ensuring sufficient data for micro calcification detection
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
Enhances breast cancer detection by identifying micro calcifications and providing high-resolution tissue characterization, improving diagnostic accuracy and reducing false positives, suitable for non-contrast enhanced breast MR imaging.
Implementation Method 1
The magnetic field B0 produces different energy levels for the individual nuclear spins in dependence on the magnetic field strength which can be excited (spin resonance) by application of an electromagnetic alternating field (RF field) of defined frequency (so-called Larmor frequency, or MR frequency)
Implementation Method 2
In order to realize spatial resolution in the body, time-varying magnetic field gradients extending along the three main axes are superposed on the uniform magnetic field B0, leading to a linear spatial dependency of the spin resonance frequency
Implementation Method 3
The transverse magnetization and its variation can be detected by means of receiving RF coils which are arranged and oriented within an examination volume of the MR device in such a manner that the variation of the magnetization is measured in the direction perpendicular to the z-axis
Data Source
AI summary
The invention relates to a method of MR imaging of an object, namely of at least a part of the female breast. The method of the invention comprises the steps of: a) subjecting the object (10) to an imaging sequence comprising RF excitation pulses and switched magnetic field gradients, wherein multiple echo signals are generated at different echo times (TE1, TE2, TE3) after each RF excitation pulse, b) acquiring the echo signals along a set of radial or spiral k-space trajectories covering a given k-space region, wherein each of the echo signals generated after an RF excitation pulse is assigned a different orientation of the radial or spiral trajectory in k-space; c) reconstructing a single echo image for each echo time (TE1, TE2, TE3) from the echo signals acquired; d) deriving high-resolution water, fat, B0 and/or apparent transverse relaxation time (T2*) maps from the single echo images using a deep learning algorithm; and d) synthesizing a high-resolution mammogram from the water, fat, B0 and/or T2* maps. Moreover, the invention relates to an MR device (1) and to a computer program for an MR device (1).