P-35 Mobile application for colorimetric analysis of paper biosensors

Presenter Status

Assistant Professor, Department of Engineering and Computer Science

Second Presenter Status

Student, Department of Engineering and Computer Science

Third Presenter Status

Professor, Department of Engineering and Computer Science

Preferred Session

Poster Session

Start Date

30-10-2015 2:00 PM

End Date

30-10-2015 3:00 PM

Presentation Abstract

Paper biosensors are a low cost, low tech diagnostic tool. A limitation of biosensors is that the sensor reaction color needs to be interpreted by a trained expert. To make the interpretation biosensors more accessible to untrained individuals, we are developing a mobile app to perform the color analysis using the on-device camera.

Photographing and analyzing the biosensor poses challenges. The app may be used in a variety of light scenarios (e.g. inside vs. outside, sunny vs. cloudy, fluorescent lighting vs. incandescent). Different lighting conditions affect the white balance of the image and complicate the sensor analysis. We propose the use of machine learning to train the app to correctly identify positive and negative readings under a variety of lighting conditions.

This is a joint project with Dr. Kwon who is developing the paper biosensors.

Acknowledgments

This project was funded by a joint 2015-2016 Faculty Research Grant (Kwon and Summerscales) from Andrews University.

This document is currently not available here.

Share

COinS
 
Oct 30th, 2:00 PM Oct 30th, 3:00 PM

P-35 Mobile application for colorimetric analysis of paper biosensors

Paper biosensors are a low cost, low tech diagnostic tool. A limitation of biosensors is that the sensor reaction color needs to be interpreted by a trained expert. To make the interpretation biosensors more accessible to untrained individuals, we are developing a mobile app to perform the color analysis using the on-device camera.

Photographing and analyzing the biosensor poses challenges. The app may be used in a variety of light scenarios (e.g. inside vs. outside, sunny vs. cloudy, fluorescent lighting vs. incandescent). Different lighting conditions affect the white balance of the image and complicate the sensor analysis. We propose the use of machine learning to train the app to correctly identify positive and negative readings under a variety of lighting conditions.

This is a joint project with Dr. Kwon who is developing the paper biosensors.