Multi-Line Recognition and Quantitative Assessment of Strip-Type Biosensors Using an Automated Image Processing Algorithm

Abstract Lateral flow assays (LFAs) are widely used point-of-care (POC) diagnostics, valued for their portability, simplicity, and rapid results. However, their conventional visual readout limits quantitative analysis and can lead to subjective interpretation and false results. In this work, we present, a novel, Python-based, automated image processing algorithm for the qualitative and quantitative analysis of multi-line LFAs capable of simultaneously detecting multiple analytes. Two multiplex formats were developed: a 4plex LFA for pathogens ( Escherichia coli (EC), Streptococcus pneumoniae (SP), Haemophilus influenzae (HI), and severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)) and a 3plex LFA for microRNAs (miR-21, miR-let-7a, and miR-155). The developed algorithm processes captured images in five stages—strip case detection, region-of-interest extraction, line segmentation, boundary detection, and density quantification—yielding both qualitative classification and quantitative analyte concentration. Performance evaluation demonstrated high specificity, no cross-reactivity, and very good quantitative accuracy compared to ImageJ-based analysis. The classification metrics of the algorithm for both multi-line LFAs provided excellent sensitivity, specificity, and overall accuracy. This integrated approach overcomes the limitations of conventional LFA interpretation, offering low-cost, portable, and scalable multi-analyte diagnostics suitable for POC applications in healthcare, food and environmental monitoring, and resource-limited settings. The platform also enables future integration with artificial intelligence (AI)-driven analytics and cloud-based remote health monitoring, supporting rapid, decentralized diagnostic capabilities.