3D Hyperspectral Biopsy Mapping for Tumor Margin Identification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods for identifying tumor margins during surgery, such as intraoperative frozen section analysis and in vivo hyperspectral imaging, are limited by accuracy, require a pathologist's presence, and fail to provide reliable information about deeper tissue layers, leading to challenges in complete tumor removal.
Innovation Solution
A method using hyperspectral imaging (HSI) to capture a biopsy sample three-dimensionally and multispectrally, generating a three-dimensional optical biopsy model with tumor-relevant segmentation values, integrated with machine learning for rapid and accurate tumor detection, including deeper tissue layers.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If intraoperative frozen section analysis is used to identify tumor margins, then the surgeon can obtain histopathological information during surgery, but the method requires a pathologist's presence, has limited reliability and accuracy, and may necessitate pausing the operation
Solution Approach 1:
The patent replaces the mechanical/frozen section-based histopathological analysis system with an optical sensing system using hyperspectral imaging. The HSI system captures spectral information from tissue samples without requiring physical sectioning, freezing, or pathologist intervention, thereby eliminating the complexity of the frozen section procedure while maintaining the ability to identify tumor margins.
Solution Approach 2:
The patent creates an optical copy or digital representation of the tissue sample through hyperspectral imaging. Instead of physically processing the tissue through frozen section, the system captures spectral data that creates a virtual model of the tissue's molecular composition, allowing tumor margin identification without altering the physical sample or requiring pathologist presence.
2Object-affected harmful factors
If in vivo hyperspectral imaging is used to detect tissue changes, then the method is non-invasive and non-ionizing, but it only measures surface layers and provides no reference point for deeper tissue layers after sample removal
Solution Approach 1:
The patent transitions from two-dimensional surface imaging to three-dimensional volumetric analysis by capturing multiple hyperspectral images at different depths or angles. This dimensional expansion allows the system to probe deeper tissue layers while maintaining the non-invasive HSI approach, providing comprehensive spatial information about both surface and deep tissue structures.
Solution Approach 2:
The patent performs preliminary spectral characterization of tissue samples before removal, capturing hyperspectral data that serves as a reference for later analysis. This preliminary action creates a spectral fingerprint or database of the tissue's molecular composition, which can then be used to identify tumor margins in the removed sample without requiring re-measurement or creating a reference point after excision.
3Productivity
If conventional color imaging methods are used to examine biopsy samples, then the method is simple and quick, but it cannot differentiate or classify samples that are indistinguishable or unreliable
Solution Approach 1:
The patent changes the measurement parameters from conventional visible light imaging to hyperspectral imaging across multiple spectral channels. This parameter expansion allows the system to detect subtle molecular differences in tissue samples that are invisible to conventional color imaging, providing precise classification and differentiation of indistinguishable samples while maintaining rapid analysis capability.
Solution Approach 2:
The patent segments the spectral information into multiple discrete wavelength channels, allowing independent analysis of different spectral features. This segmentation enables the system to identify and classify subtle differences in tissue composition by examining specific spectral signatures, thereby differentiating samples that appear identical in conventional color imaging.
4Loss of information
If hyperspectral imaging is applied to create three-dimensional models of biopsy samples, then the method provides complete spatial information and geometric impression, but it requires complex data processing and machine learning algorithms
Solution Approach 1:
The patent implements self-service through automated machine learning algorithms that automatically process and interpret the hyperspectral data without requiring manual analysis. The system performs its own classification, segmentation, and three-dimensional model generation through integrated algorithms, eliminating the need for manual intervention while managing the computational complexity through automation.
Solution Approach 2:
The patent introduces machine learning algorithms as an intermediary between the raw hyperspectral data and the final diagnostic output. These algorithms serve as a mediator that processes the complex spectral information, extracts relevant features, and generates interpretable results, thereby managing the computational complexity and making the system more user-friendly while maintaining complete spatial information.
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
Enables rapid, accurate, and reliable identification of tumor margins, reducing the need for histological diagnostics and facilitating complete tumor resection by providing a geometric, three-dimensional impression of the biopsy sample, aligning with the in vivo situation.
Implementation Method 1
a first camera coordinate x and a second camera coordinate y, which span a camera image plane perpendicular to an optical axis Z of the HSI camera 11
Implementation Method 2
From the majority of the reflective intensity images, a three-dimensional optical biopsy model is generated using a photogrammetric reconstruction method
Data Source
Figure 1
Figure 2
Figure 3
AI summary
The invention relates to a method for assigning at least one tumor-relevant segmentation value to each surface point of a biopsy sample (P). The surface of the sample is captured at least partially three-dimensionally and multispectrally using a hyperspectral imaging (HSI) camera (11). Several hypercubes (H) are captured in separate camera image planes. A reflective intensity image is acquired for each hypercube (H) or derived from the hypercube (H). A three-dimensional optical biopsy model (M1) is generated from the majority of the reflective intensity images using a photogrammetric reconstruction method. Each hypercube (H) is transformed into a segmentation map (S) by means of an evaluation unit (12) trained with a machine learning method. This map assigns a tumor-relevant segmentation value to each pixel of the camera image plane of the hypercube (H).The segmentation values of the segmentation maps (S) are transferred to the respective surface point of the optical biopsy model (M1) that corresponds to the respective pixel of the camera image plane according to the photogrammetric reconstruction. The invention further relates to a device for carrying out this method.