Implementing Mass Testing Approach with Computer Vision Techniques

In medical laboratories, medical tests are taken to be analyzed for providing better treatment to patients, screening diseases in the population, and for further research on diseases. During the times of the rise of epidemics and pandemics, a mass testing approach for screening the spread of diseases in the population is used. It takes several hours for health workers to manually record the results in the database as presented during the outbreak of the COVID-19 pandemic, and this caused late notification to the infected which resulted in further spread. Based on object recognition with computer vision techniques, the OCR engine, and advanced data analysis techniques, an efficient novel approach to develop a novel deep learning model to classify images while extracting insightful information from them for further manipulation and use has been successfully achieved. The developed model contributes to designing the Bioline RDT scanner machine. The machine is designed to scan different rapid diagnosis test (RDT) types for invalid, negative, or positive results for single or many patients, and extract the corresponding patients’ IDs at a time. As the RDT types are not predefined during the development of the model, all the existing and future RDT to come can be monitored by the machine.