Diffuse Optical Tomography Lesion Reconstruction Using Pseudoinverse Regularization
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current diagnostic imaging techniques for breast cancer face challenges in accurately differentiating between malignant and benign lesions due to overlapping appearances, leading to a high number of unnecessary biopsies and limited success in improving this situation, with diffuse optical tomography (DOT) systems facing issues in lesion localization and light quantification due to intensive light scattering and underdetermined reconstruction problems.
Innovation Solution
A system and method that utilize a diffuse optical tomography device to acquire functional data from breast lesions and healthy tissue, transforming the data into perturbation data to generate optimized functional images by applying a pseudoinverse matrix and regularized optimization methods, which provide hemoglobin concentration maps to improve lesion characterization and reduce unnecessary biopsies.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional imaging techniques are used for breast cancer diagnosis, then screening coverage is achieved, but differentiation between malignant and benign lesions is poor leading to high number of unnecessary biopsies
Solution Approach 1:
The patent applies functional imaging to detect hemoglobin concentration and oxygen saturation, which provide functional characteristics (vascularity, metabolism) that differentiate malignant from benign lesions. This functional information acts as a new 'color' or dimension of tissue characterization that conventional structural imaging cannot provide, enabling better lesion classification and reducing unnecessary biopsies
Solution Approach 2:
The patent combines structural imaging data with functional imaging data (hemoglobin concentration, oxygen saturation) to create a composite assessment of lesions. This multi-parametric approach integrates multiple types of information to achieve more accurate lesion characterization than either imaging modality alone
2Measurement precision
If DOT is used to improve lesion localization and light quantification, then functional imaging capability is enhanced, but intensive light scattering in tissue causes reconstruction challenges
Solution Approach 1:
The patent applies preliminary action by using structural imaging to identify and localize lesions before performing functional DOT imaging. This pre-localization step allows the DOT system to focus its functional assessment on specific regions of interest, reducing the overall complexity of the reconstruction problem by limiting the spatial domain that requires detailed functional mapping
Solution Approach 2:
The patent uses structural imaging as an intermediary that guides and constrains the DOT functional reconstruction. The structural image provides anatomical priors and region-of-interest definitions that serve as mediators between the raw optical measurements and the final functional maps, simplifying the inverse problem by incorporating anatomical constraints
3Measurement precision
If iterative optimization methods are applied to generate functional images, then image quality is improved, but computation time increases
Solution Approach 1:
The patent performs preliminary lesion localization using structural imaging before executing the computationally intensive DOT reconstruction. This preliminary step allows the system to restrict the functional reconstruction to small, predefined regions of interest containing lesions, dramatically reducing the number of voxels and computational operations required while maintaining functional image accuracy
Solution Approach 2:
The patent segments the breast tissue into multiple regions, with detailed functional reconstruction applied only to lesion-containing regions identified by structural imaging. This spatial segmentation strategy applies high computational effort only where necessary (in lesion regions) while using lower-resolution or no functional imaging in benign tissue regions, optimizing the trade-off between image quality and computation time
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
The approach enhances the accuracy of vasculature distribution and reduces computation time, allowing for more precise differentiation between malignant and benign lesions, thereby decreasing the number of unnecessary biopsies and improving cancer detection sensitivity while maintaining low-risk assessment.
Implementation Method 1
Due to intensive light scattering in tissue, lesion localization and light quantification accuracy is an ongoing challenge
Data Source
AI summary
A system and method for generating at least one optimized functional images of a lesion region of a subject is provided. The system includes a diffuse optical tomography (DOT) device, and a computing device. The DOT device is configured to acquire lesion functional data of an imaging volume including the lesion region of the breast and reference functional data from a corresponding imaging volume including healthy tissue within a contralateral breast. The computing device is programmed to generate at least one functional image of the breast by reconstructing the functional data at the plurality of regions including the lesion region and the surrounding adjacent background region. The functional images may be reconstructed by an optimization method regularized a preliminary estimate generated by applying a truncated pseudoinverse matrix of a weight matrix to the functional data. The optimized functional images relate to levels of hemoglobin at the voxels.