Motion-Corrected Fetal Brain Diffusion Tensor Imaging

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current diffusion tensor imaging techniques face challenges in capturing geometrically consistent data from moving objects, particularly in fetal brain imaging, due to unpredictable motion, which corrupts the spatial correspondence between images and requires prolonged maternal breath-hold, making it difficult to acquire reliable quantitative parameters.

Innovation Solution

A method that uses dynamic EPI scanning with multiple overlapping slices and image registration to align scattered data, reconstructing a 3D b=0 volume and diffusion tensor matrix on a regular Cartesian lattice, accounting for motion and ensuring accurate diffusion tensor estimation despite significant motion.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional diffusion tensor imaging is used to capture quantitative parameters, then measurement precision is improved, but the acquisition time increases and motion corruption occurs

Engineering Contradiction:
Improvequantitative parameter measurementVSAvoidacquisition time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The imaging volume is divided into multiple thin slices that are acquired sequentially. Each slice is captured in a motion-frozen state during maternal breath-hold, allowing multiple slices to be acquired within a single breath-hold period. This segmentation enables comprehensive 3D coverage without extending the total breath-hold time.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent transitions from conventional single-shot 3D DTI to multi-slice 2D imaging with subsequent volumetric reconstruction. By acquiring multiple 2D slices at different positions and reconstructing them into a 3D volume, the method achieves 3D isotropic resolution while maintaining compatibility with breath-hold time constraints.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

2Measurement precision

If multiple diffusion weighted images are acquired for DTI, then measurement precision is improved, but the acquisition time increases requiring prolonged maternal breath-hold

Engineering Contradiction:
Improvediffusion tensor estimationVSAvoidmaternal breath-hold duration
Core Design Contradiction:
Measurement precisionVSDuration of action of moving object

Solution Approach 1:

The complete DTI protocol is segmented into multiple slices, each acquired during a single maternal breath-hold. This allows the necessary multiple diffusion-weighted images per slice to be obtained without requiring an unmanageably long continuous breath-hold, as each slice represents a discrete sampling opportunity.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent acquires more diffusion-weighted images than the minimum 6 directions required for DTI calculation. By acquiring additional slices and potentially redundant diffusion directions, the method ensures sufficient data quality despite the constraints on total acquisition time and maternal breath-hold duration.

Inventive Principle:
Principle #16Partial or excessive action

3Device complexity

If standard slice acquisition is used, then device complexity is reduced, but motion corruption of spatial correspondence occurs

Engineering Contradiction:
Improveimaging sequenceVSAvoidspatial correspondence
Core Design Contradiction:
Device complexityVSReliability

Solution Approach 1:

Image registration is performed as a preliminary step to establish the spatial correspondence between multiple slices before diffusion tensor calculation. This pre-alignment ensures that all slices are correctly positioned relative to each other in 3D space, correcting for any motion that occurred between slice acquisitions and preventing motion corruption in the final DTI parameters.

Inventive Principle:
Principle #10Preliminary action

4Productivity

If single shot techniques are used for fetal imaging, then acquisition time is reduced, but geometric consistency is lost

Engineering Contradiction:
Improveacquisition speedVSAvoidgeometric consistency
Core Design Contradiction:
ProductivityVSStability of the object's composition

Solution Approach 1:

The imaging approach is segmented into multiple thin slices acquired sequentially during free-breathing conditions, rather than attempting to capture the entire volume in a single shot. This segmentation allows each slice to be acquired quickly while maintaining the ability to reconstruct a geometrically consistent 3D volume through registration and interpolation.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent uses 2D slice acquisition with subsequent 3D volumetric reconstruction to achieve geometric consistency. By treating the data as a set of 2D slices that are registered and interpolated into 3D space, the method recovers geometric consistency that would be lost in conventional single-shot approaches, while maintaining fast acquisition suitable for fetal imaging.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

Data Source

PatentEP2145200B1Diffusion tensor imaging of moving objects
Publication Date: 2015.09.16 IP2IPO INNOVATIONS LTD
  • EP2145200B1 patent drawingFigure 1(a)~1(c)
  • EP2145200B1 patent drawingFigure 2(a)~2(c)
  • EP2145200B1 patent drawingFigure 3(a)~3(c)

AI summary

A method of generating an image data set, describing an object, as the object moves in a scanning space, comprises: a) performing a plurality of scans of the scanning space each scan being arranged to generate a set of samples, each sample including a sample value of at least one parameter associated with a respective point in the scanning space, the plurality of scans generating sample sets relating to a plurality of different parameters; b) determining a mapping for each of the samples from the scanning space into an object space which is fixed relative to the object; and c) determining object space values for each of the parameters at each of a plurality of points in the object space.