Background
The global COVID-19 pandemic has significantly expedited the adoption of lateral flow assays. While interpreting COVID-19 test results is typically straightforward, diagnosing complex infectious and chronic diseases, such as malaria, cardiovascular conditions, and cancer, using lateral flow assays presents greater challenges. These tests often produce multiplexed results with multiple test lines that require specialized expertise for accurate interpretation. Moreover, suboptimal human interpretation, particularly in cases involving multiple or faint test lines, can further compromise diagnostic accuracy and lead to a reduction in both sensitivity and specificity.
The RSE team collaborated with Professor Dame Molly Stevens and Dr. Ho-Cheung Ng from the Stevens Group at Imperial College London and the University of Oxford to develop an AI-powered application that leverages smartphone cameras to analyse lateral flow tests, providing highly accurate quantification of results. In cases of ambiguous predictions, the system transmits the data to a central server for further analysis, where experts validate and ensure the correct interpretation of the results.
Our Contribution
The project resulted in a robust web application built on Django, containerized for portability, and a mobile application developed using Flutter, enabling deployment on both iOS and Android devices. The platform has been designed to be extensible, allowing the Stevens Group to seamlessly integrate and evaluate new AI models as they develop additional lateral flow assay types.
Testimonials
Dr Ho-Cheung Ng, project leader and Postdoctoral Researcher in The Stevens Group:
“The RSE team played a pivotal role in advancing our analysis of lateral flow assays by developing a sophisticated smartphone application. Their professionalism and adaptability in integrating various machine learning models were invaluable. Additionally, they took the time to educate our team on software best practices and research software design, which was greatly appreciated.”