We found a match
Your institution may have access to this item. Find your institution then sign in to continue.
- Title
PuriScan: A Low-Cost Microscope for Pathogenic Detection using Machine Learning and Convolutional Neural Networks.
- Authors
Cheng, George
- Abstract
2.1 billion people worldwide lack access to potable water; roughly 250 million people developed malaria in 2020 alone, and an overwhelming majority of the cases were in underdeveloped African countries. This research aims to remove socioeconomic disparities in drinking water by introducing a novel, cost-effective, and relatively accessible method for pathogen detection, generally focusing on widespread and hazardous bacteria (e.g., Escherichia coli, Salmonella) and protozoa (e.g., Amoeba, Paramecium, Euglena. By implementing a machine learning (M.L.) algorithm for pathogen identification and consolidating the data into a cloud database to improve the convolutional neural networks (CNN), PuriScan, our app, and product can accurately and rapidly assess the contents of water samples in less than 10 seconds, costing only 20 cents each. Furthermore, with confidence levels of 99% through thousands of selected database images in both pathogenic presence and identification, PuriScan can precisely pinpoint the location of dirty water through a GPS-coordinate system and “map” the feature. Integrating PuriScan’s pure engineering with aspects of computerized intelligence can significantly reduce the logistics of traditional water characterization.
- Subjects
DRINKING water; SOCIOECONOMIC disparities in health; MACHINE learning; CONVOLUTIONAL neural networks; PARAMECIUM
- Publication
International Journal of High School Research, 2023, Vol 5, Issue 2, p42
- ISSN
2642-1046
- Publication type
Article
- DOI
10.36838/v5i2.9