Image Data Conversion via Regression Modeling
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current radiological imaging methods face challenges in converting image data from inexpensive or position-specific imaging systems into forms suitable for medical interventions, often resulting in suboptimal surgical procedures due to positional discrepancies and the limitations of existing biomechanical models that cannot be easily transferred across different body areas.
Innovation Solution
A method and device that use a data-driven regression approach to determine a set of target shape parameters from input pixels, allowing for the conversion of image data from one form to another, such as from a prone to a supine position, without the need for additional recordings, using a target shape model that can accurately represent image data with a reduced number of parameters, thus enabling efficient and cost-effective image data conversion across different imaging modalities and positions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of manufacture
If image data is acquired using cost-effective imaging systems or in specific positions, then examination costs are reduced and routine examinations become feasible, but the image data does not correspond to the body part's state during subsequent medical procedures
Solution Approach 1:
The patent creates a virtual copy of the image data by generating a three-dimensional model from the two-dimensional image data. This virtual model allows medical personnel to visualize and plan procedures as if they were viewing the actual three-dimensional structure, enabling accurate intervention planning without requiring expensive additional imaging or repositioning the patient.
Solution Approach 2:
The patent transforms image data between different parameter spaces by converting two-dimensional image coordinates into three-dimensional spatial coordinates. This parameter transformation allows the same image data to be interpreted in multiple dimensional contexts, enabling accurate procedure planning from routine examination images.
2Measurement precision
If additional medical equipment is used to fix the body area during image acquisition, then image quality is standardized and movement is avoided, but the equipment hinders subsequent medical procedures
Solution Approach 1:
The patent creates a virtual representation of the body part's three-dimensional structure from two-dimensional images, eliminating the need for physical manipulation or special positioning equipment during procedures. The virtual model preserves anatomical accuracy while allowing unrestricted access during medical interventions.
3Measurement precision
If biomechanical models with specific parameters are used for mapping regions, then mapping accuracy is improved for specific body areas, but the models cannot be easily transferred to other body regions
Solution Approach 1:
The patent develops a universal three-dimensional modeling approach that can be applied to any body part regardless of its specific biomechanical properties. The method uses general image processing and spatial transformation techniques that work across different anatomical regions, eliminating the need for region-specific biomechanical models while maintaining accuracy.
Data Source
Figure 1~2
Figure 3
Figure 4
AI summary
The method involves determining a set of defined multiple input pixels (X1-Xk) in image data of a first form (1). A set of targets is prognostically determined from parameters of a target form model with defined target form parameters (Y1-Yl) by data-driven regression process of multiple input pixels. A second form (2) of image data is determined from the set of target form parameters, which describes a deviation from a standardized geometric target form model. A standardized geometric target form model is determined from given radiological training image data sets. Independent claims are also included for the following: (1) a method for determining a geometric position of a number of target objects (2) an image processing workstation (3) a target object determination device (4) an imaging device (5) a computer program product comprising a set of instructions to perform a method for determining image data of an examination object.