Behind Density Lines: Machine Learning and Citizen Scientists in Quantifying Scanning Electron Microscopy Images

Monét Roberts1
Yinlin Chen2
Tanner Upthegrove3

1Department of Biomedical Engineering and Mechanics, 2University Libraries, 3Institute for Creativity, Arts, and Technology
Virginia Tech, Blacksburg, VA, USA

Amid the pandemic, screen interactions have surged by 60-80%, accentuating digital eye strain, attention deficits, and posture issues. This study addresses a key challenge in image analysis post-acquisition, especially in scanning electron microscopy (SEM) critical for health and material sciences applications. For instance, surface morphologies in cancer research and fiber quantification in biomaterials demand precise analysis, often hindered by time constraints and human error. This project proposes a machine learning workflow to accurately quantify SEM images, leveraging previously quantified SEM images to create a high-quality training dataset. Concurrently, a community engagement pipeline is devised to involve citizen scientists in SEM analysis, beginning with a pilot study and extending to interactive sessions at the Virginia Tech Science Festival. This dual approach aims to bolster SEM analysis accuracy and foster a reciprocal educational ecosystem, setting a strong foundation for future machine learning applications and community-driven scientific explorations.

Image 1 Alt Text    Image 2 Alt Text    Image 2 Alt Text

Virginia Tech Science Festival

Behind Density Lines: An Embodied Scanning Electron Microscopy Game

Scanning Electron Microscopes can show us what is happening on an incredibly small scale! Scientists are still learning how to use Machine Learning and other new technologies to identify and quantify what is in these images. This exhibit empowers anyone to be a scientist by playing a drawing game to quantify membrane protrusions on Mucin-1 overexpressing non-tumorigenic breast cells. Versions of the game will be available on tablets, as well embodied by your movements in the Cube.

Saturday, November 11, 2023
10 AM - 3 PM
Moss Arts Center
Free and open to the public

CS4624: AI & Citizen Science in SEM Analysis
Student Team:
Anthony Nguyen John Siegel
Alex Lin Luke DiGiovanna
GitHub Report/Presentation

This project is supported by University Libraries Collaborative Research Grant.