AR/VR-Based Whole Slide Visualization for 3D Tissue Depth
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Pathology slides presented in two dimensions fail to accurately represent the irregular, three-dimensional shapes and dimensions of biological structures, making it difficult for pathologists to assess and visualize the depth of tissue samples.
Innovation Solution
Utilizing augmented reality (AR) and virtual reality (VR) technologies to generate navigable three-dimensional images of tissue samples by integrating machine-learning models that analyze whole slide images (WSIs), allowing pathologists to interactively navigate and analyze the images using hand or eye gestures.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If pathology slides are presented as two-dimensional images, then the presentation is simple and easy to process, but the ability to visualize three-dimensional tissue depth and structure is lost
Solution Approach 1:
The patent transforms two-dimensional whole slide images into a three-dimensional navigable space by stacking images along the Z-axis (depth dimension). This allows pathologists to perceive tissue depth and three-dimensional structure while maintaining the simplicity of image-based presentation. The system creates a virtual microscope environment where users can navigate through multiple focal planes to explore volumetric tissue architecture.
2Device complexity
If multiple whole slide images are presented independently, then the processing complexity is reduced, but the ability to assess three-dimensional biological structures is difficult
Solution Approach 1:
The patent merges multiple independent whole slide images into a single integrated three-dimensional navigable space. By combining images from different focal planes and stacking them along the Z-axis, the system creates a unified volumetric representation that enables accurate assessment of three-dimensional biological structures while maintaining manageable system complexity through automated image registration and navigation algorithms.
3Ease of manufacture
If two-dimensional piecemeal presentation is used, then the system is easier to implement, but the visualization of irregular three-dimensional shapes is difficult
Solution Approach 1:
The patent addresses the limitation of two-dimensional presentation by adding the depth dimension (Z-axis) to the image stack. This creates a three-dimensional navigable space that naturally accommodates irregular three-dimensional biological shapes, enabling accurate visualization and assessment while maintaining ease of implementation through automated image processing and navigation frameworks.
Data Source
AI summary
A computer-implemented method for generating a navigable three-dimensional image of a tissue sample using augmented reality or virtual reality (AR/VR), the method comprising: receiving, by one or more processors, a plurality of medical images, the plurality of medical images being associated with a whole slide image (WSI) the tissue sample; determining by one or more processors, a layer level of each medical image of the plurality of medical images; determining, by one or more processors, for each medical image of the plurality of medical images, one or more regions of tissue and one or more regions of non-tissue; stacking, by one or more processors, each medical image with one or more of the plurality of medical images based on the layer level to generate a three-dimensional rendering for display in an AR/VR environment; and outputting, by one or more processors, the AR/VR environment as seen in a display.