Case Study: Smartphone Lateral Flow Reader
Introduction
We have developed Lateral Flow Reader software for over 7 years, our sister company MicroLab Devices was one of the first companies to file a patent in this field, back in 2012. In this time we have worked with SME's and large multi-nationals to develop various Lateral Flow Device based solutions using both smartphone based technology and desktop based instrumentation.
This case study looks at a smartphone based application we developed with our partner Soma Bioscience. The application was designed to work with their saliva based Cortisol and IgA lateral flow device.
Key Points
Completed System
Technologies
Customer
External Links & Publications
Lateral Flow Detection & Image Processing
The first step is to design a detection routine for detecting the specific Lateral Flow Device (LFD). Working with the smartphone camera we are able to design real-time detection algorithms which run on Android and iOS. These algorithms provide an easy to use and reliable capture method for end-users.
Once the LFD is detected key features are extracted for analysis. This uses a range of image processing techniques. The detection and extraction process is the most challenging since it relies on many external factors including lighting, smartphone type and operator handling.
Lateral Flow Analysis & Workflow
After the key features are extracted they are processed and converted into a meaningful output or calibrated units or custom algorithm representing some normalised scale. This is just part of the overall software workflow. We designed a complete custom workflow, providing the customer with instructions, protocol videos, batch management, they can create PDF reports which can be emailed or the data stored in the cloud or on Laboratory Information Systems.
Data Portal & Back-end
The data and results are stored in the cloud, using AWS. The portal allows users to manage their own credentials and review their stored data. Furthermore, our customer (the LFD manufacturer) is able to log in and add batch and calibration data and modify algorithms all in real-time.
