Tomographic Image Reconstruction via Dual Inversion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional hybrid imaging techniques face limitations in reconstructing fluorescence distributions within objects, particularly due to the reliance on incomplete anatomical information and limited image quality when combining fluorescence tomography with other modalities.
Innovation Solution
A system and method that utilize a computing device to combine first and second tomography data, where the computing device calculates a forward model using second tomography data to perform a first inversion, generating a parameter set which is then used for a second inversion to obtain an improved object image, incorporating anatomical information from modalities like X-ray CT or MRI to enhance fluorescence distribution recovery.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If conventional hybrid imaging techniques combine fluorescence tomography with other modalities, then anatomical information is provided, but image quality and accuracy of fluorescence distribution reconstruction deteriorates due to incomplete anatomical information
Solution Approach 1:
The patent applies segmentation by dividing the reconstruction process into two distinct inversion steps: a first inversion that uses only optical tomography data, and a second inversion that integrates anatomical information from secondary modalities. This segmentation allows each inversion to specialize - the first captures fluorescence distribution without anatomical constraints, while the second refines results using anatomical priors, thereby resolving the contradiction between having anatomical information and maintaining reconstruction accuracy
Solution Approach 2:
The patent introduces an intermediary approach by using the first inversion result as a baseline that is then refined by the second inversion. The first inversion acts as an intermediary step that preserves pure fluorescence information, which then serves as input for the second inversion that incorporates anatomical data. This intermediary process prevents direct contamination of fluorescence reconstruction by potentially incomplete anatomical information
2Device complexity
If a single inversion process is used to combine tomography data, then processing is simpler, but image quality and fluorescence distribution recovery deteriorates
Solution Approach 1:
The patent segments the inversion process into two distinct stages: first inversion that processes optical tomography data independently, and second inversion that integrates anatomical information from secondary modalities. This segmentation increases processing complexity but dramatically improves fluorescence distribution recovery accuracy by allowing each inversion to optimize for its specific data type and purpose
3Loss of information
If anatomical information from secondary modalities is incorporated, then comprehensive object image is obtained, but processing time and computational effort increases
Solution Approach 1:
The patent applies preliminary action by performing the first inversion using only optical tomography data before incorporating anatomical information from secondary modalities. This preliminary step establishes a baseline fluorescence distribution that can then be refined by anatomical constraints in the second inversion, reducing the computational burden of processing all data simultaneously while still achieving comprehensive object image reconstruction
Data Source
Figure 1~3
Figure 2
Figure 4
AI summary
A system for creating an object image of an object under investigation, comprises a computing device being configured for creating the object image by using first tomography data of the object provided by a first tomography system and second tomography data of the object provided by a second tomography system, wherein the computing device is configured for calculating a first forward model describing the first tomography data using the second tomography data, performing a first inversion of the first forward model, calculating a parameter set using the first inversion, and performing a second inversion using the parameter set to obtain the object image to be created. Furthermore, a method of imaging an object under investigation is described, wherein the system for creating an object image is used.