Towards Human in the Loop Analysis of Complex Point Clouds: Advanced Visualizations, Quantifications, and...

Area of Science:

• * Computational biology and medical imaging.
• * Data visualization and analysis.
• * Virtual Reality (VR) applications in science.

Background:

• * Biological and medical research generates high-dimensional point cloud data from experiments like microscopy and medical imaging.
• * Analyzing complex, noisy point cloud data presents ergonomic and computational challenges for user interaction and real-time processing.
• * Existing visualization software often lacks flexibility in data manipulation and analysis, and VR experiences can be hindered by computational demands.

Purpose of the Study:

• * To introduce new libraries for enhancing interaction and human-in-the-loop analysis of point cloud data within a virtual reality environment.
• * To integrate these libraries into the open-source platform Genuage, improving user experience and analytical flexibility.
• * To enable simultaneous, fluid manipulation and analysis of complex data with high refresh rates.

Main Methods:

• * Development of a communication toolbox to enhance user experience and flexibility in VR data interaction.
• * Implementation of a mapping toolbox for overlaying physical properties onto 3D meshes while using a dedicated point cloud shader.
• * Creation of a programmable video capture tool for VR and desktop modes to facilitate data dissemination.
• * Establishment of protocols for simultaneous analysis and fluid data manipulation at high refresh rates.

Main Results:

• * Enhanced user interaction and flexibility in analyzing complex point cloud data within VR.
• * Successful integration of new libraries into the Genuage platform, supporting advanced visualization and analysis.
• * Demonstrated real-time inference of random walk properties using a pre-trained Graph Neural Network within the VR environment.
• * Achieved fluid manipulation and analysis of data with a high refresh rate, overcoming previous computational limitations.

Conclusions:

• * The developed libraries significantly improve the analysis of high-dimensional point cloud data in VR for biological and medical research.
• * The Genuage platform, enhanced with these tools, offers a more flexible and interactive environment for scientific data exploration.
• * Real-time analysis and manipulation are feasible, paving the way for more efficient and intuitive scientific discovery.