3D Anatomy Scan Reformatting With Landmark Mask Alignment

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current clinical imaging techniques, such as MRI and CT, require repetitive scanning and manual reformatting of 3D data to obtain contiguous views of anatomical landmarks, which is time-consuming and sensitive to technician skill, leading to potential errors in diagnosis.

Innovation Solution

A system utilizing a pre-trained neural network to generate deep-learning estimated scan prescription masks, allowing for automated alignment and reformatting of 3D image data to provide continuous views of anatomical landmarks from a single scan, reducing scan time and improving accuracy.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If repetitive scanning from different acquisition planes is performed to obtain contiguous views of anatomical landmarks, then the quality and completeness of anatomical imaging is improved, but the scan time is significantly increased

Engineering Contradiction:
Improveanatomical imaging qualityVSAvoidscan time
Core Design Contradiction:
Manufacturing precisionVSLoss of time

Solution Approach 1:

The patent performs preliminary actions by acquiring a single 3D data stack with comprehensive anatomical coverage before reformatting. The 3D acquisition is performed in advance with sufficient field of view to include all desired anatomical landmarks, eliminating the need for repetitive scanning. The reformatting process then retrospectively generates the required views from this pre-acquired 3D data.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent transitions from multiple 2D acquisition planes to a single 3D volumetric acquisition. By acquiring data in three dimensions with isotropic or near-isotropic resolution, the system can retrospectively generate contiguous views of any anatomical landmark from the 3D data stack, eliminating the need for repetitive 2D scanning from different planes.

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

2Adaptability or versatility

If manual reformatting of 3D data is performed by technicians to obtain views of desired landmarks, then the flexibility to obtain specific anatomical views is improved, but the process becomes time-consuming and sensitive to technician skill

Engineering Contradiction:
Improveflexibility in obtaining anatomical viewsVSAvoidreformatting time
Core Design Contradiction:
Adaptability or versatilityVSLoss of time

Solution Approach 1:

The patent implements self-service by using automated algorithms and software to perform the reformatting process without requiring technician intervention. The system automatically identifies anatomical landmarks, determines optimal reformatting planes, and generates contiguous views of desired landmarks from the 3D data stack, making the process independent of technician skill and significantly reducing reformatting time.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent replaces the manual mechanical process of technician-based reformatting with an automated computational system. Instead of technicians manually manipulating imaging consoles to achieve desired views, the system uses automated image processing algorithms to retrospectively generate reformatted views from the 3D data, eliminating the time-consuming manual process while maintaining flexibility.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

3Manufacturing precision

If higher resolution 3D data stack is acquired to improve anatomical detail, then the quality of reformatted views is improved, but the data processing and reformatting complexity increases

Engineering Contradiction:
Improveanatomical detail qualityVSAvoiddata processing complexity
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent performs preliminary actions by acquiring the high-resolution 3D data stack with sufficient field of view to include all desired anatomical landmarks in a single acquisition. This pre-acquisition of comprehensive 3D data with isotropic resolution eliminates the need for subsequent repetitive scanning and simplifies the reformatting process, as all necessary anatomical information is already captured in the initial 3D acquisition.

Inventive Principle:
Principle #10Preliminary action

4Measurement precision

If graphical prescription or reformatting parameters are manually adjusted to achieve accurate anatomical alignment, then the accuracy of anatomical landmark visualization is improved, but the process becomes sensitive to small errors and requires repeated scans

Engineering Contradiction:
Improveanatomical landmark alignment accuracyVSAvoiddiagnosis reliability
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent replaces manual graphical prescription and reformatting parameter adjustment with automated computational algorithms. The system automatically identifies anatomical landmarks, determines optimal reformatting planes and orientations, and generates contiguous views with high precision. This automated approach eliminates human error and sensitivity to small parameter adjustments, significantly improving the reliability of anatomical landmark visualization and reducing the need for repeated scans.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentEP4086852A1Generating reformatted views of a three-dimensional anatomy scan using deep-learning estimated scan prescription masks
Publication Date: 2022.11.09 GE PRECISION HEALTHCARE LLC
  • EP4086852A1 patent drawingFigure 1
  • EP4086852A1 patent drawingFigure 2
  • EP4086852A1 patent drawingFigure 3

AI summary

Techniques are described for generating reformatted views of a three-dimensional (3D) anatomy scan using deep-learning estimated scan prescription masks. According to an embodiment, a system is provided that comprises a memory that stores computer executable components, and a processor that executes the computer executable components stored in the memory. The computer executable components comprise a mask generation component that employs a pre-trained neural network model to generate masks for different anatomical landmarks depicted in one or more calibration images captured of an anatomical region of a patient. The computer executable components further comprise a reformatting component that reformats 3D image data captured of the anatomical region of the patient using the masks to generate different representations of the 3D image data that correspond to the different anatomical landmarks.