Generalized Spherical Deconvolution for Accelerated Diffusion MRI
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current diffusion MRI techniques require extended scan times to acquire high-resolution images, leading to patient discomfort and potential image quality issues due to movement during acquisitions, as they need to sample the entire 3D q-space, which is time-consuming and results in undersampled data with artifacts.
Innovation Solution
The method combines compressed sensing and generalized spherical deconvolution to process undersampled q-space data, transforming it into a fully sampled dataset, thereby reducing acquisition times without compromising image quality by denoising and reconstructing spatially resolved fiber orientation distributions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional diffusion MRI techniques sample the entire 3D q-space to achieve high-resolution images, then image resolution is improved, but scan time is extended
Solution Approach 1:
The patent applies partial action by acquiring only a subset of q-space data points rather than sampling the entire 3D q-space. Undersampled data is collected and then reconstructed using compressed sensing and spherical deconvolution algorithms to recover the complete fiber orientation information, thereby reducing scan time while maintaining image resolution
Solution Approach 2:
The patent substitutes the mechanical data acquisition process with computational reconstruction. Instead of physically acquiring all q-space points through extended scanning, the system uses mathematical algorithms (compressed sensing and spherical deconvolution) to reconstruct the full dataset from undersampled measurements, replacing time-consuming mechanical sampling with efficient computational processing
2Measurement precision
If extended scan times are used to acquire acceptable diffusion data, then image quality is improved, but patient movement increases causing artifacts
Solution Approach 1:
The patent reduces the acquisition time by collecting only partial q-space data, which minimizes the duration patients must remain stationary. The compressed sensing reconstruction algorithm then recovers the complete image quality from this reduced dataset, thereby reducing movement artifacts while maintaining diagnostic image quality
Solution Approach 2:
The patent performs preliminary computational processing (compressed sensing reconstruction and spherical deconvolution) on undersampled data to predict and recover the complete q-space information before final image reconstruction. This preliminary computational action allows the system to achieve full image quality without requiring patients to endure extended scan times that would increase movement artifacts
3Loss of time
If q-space data is undersampled to reduce scan time, then acquisition time is reduced, but data quality deteriorates with artifacts
Solution Approach 1:
The patent replaces the mechanical data collection process with computational reconstruction. Compressed sensing algorithms reconstruct the complete q-space dataset from undersampled measurements, and spherical deconvolution further processes this reconstructed data to recover accurate fiber orientation distributions, thereby eliminating quality deterioration despite time reduction
Solution Approach 2:
The patent introduces computational algorithms (compressed sensing and spherical deconvolution) as intermediaries between the undersampled raw data and the final high-quality images. These intermediary processing steps recover the missing information and eliminate artifacts, bridging the gap between reduced acquisition time and maintained data quality
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
This approach accelerates diffusion MRI acquisition times while maintaining or improving image resolution and quality, allowing for better visualization of fiber crossings and tractography with reduced patient discomfort and movement artifacts.
Implementation Method 1
a primary field magnet that may place gyromagnetic nuclei within a patient into an equilibrium magnetization
Implementation Method 2
a plurality of gradient field coils that may encode positional information into the gyromagnetic nuclei
Implementation Method 3
a radiofrequency (RF) transmit coil that may perturb the gyromagnetic nuclei away from their equilibrium magnetization
Implementation Method 4
a plurality of RF receiving coils configured to receive MR signals from the gyromagnetic nuclei as they relax to their equilibrium magnetization
Implementation Method 5
The movement of the water molecules may be characterized as incoherent motion, which results from diffusion processes
Data Source
AI summary
A magnetic resonance imaging method includes generating spatially resolved fiber orientation distributions (FODs) from magnetic resonance signals acquired from a patient tissue using a plurality of diffusion encodings, each acquired magnetic resonance signal corresponding to one of the diffusion encodings and being representative of a three-dimensional distribution of displacement of magnetic spins of gyromagnetic nuclei present in each imaging voxel. Generating the spatially resolved FODs includes performing generalized spherical deconvolution using the acquired magnetic resonance signals and a modeled tissue response matrix (TRM) to reconstruct the spatially resolved FODs. The method also includes using the spatially resolved FODs to generate a representation of fibrous tissue within the patient tissue.