Evaluation of Fluorescence Detection Algorithms for Efficient ROI Setting in Low-Cost Real-Time PCR Systems.

This study proposes a region of interest (ROI) setting method to improve the accuracy and efficiency of fluorescence detection in a compact real-time multiplex fluorescence PCR system. Conventional commercial real-time PCR systems are limited in point-of-care (POC) environments due to their high cost and complex optical structures. To address this issue, we developed a low-cost, compact system using an open-platform camera and a Fresnel lens. However, in such a simply structured system, variations between the wells of the polymerase chain reaction (PCR) plate may affect the accuracy of fluorescence detection. In this study, after capturing images with a CMOS camera, we propose two ROI image processing algorithms. The proposed algorithms reliably extract fluorescence signals and compare ROI deviations caused by variations between wells to determine whether physical correction is necessary. To validate the system, we performed comparative analysis of real-time DNA amplification images and fluorescence dye images collected over multiple periods. Based on evaluations using manual detection as a reference, it was confirmed that even a simple algorithm can achieve stable fluorescence detection while minimizing ROI distortion. This study presents an efficient method for enhancing the accuracy of quantitative fluorescence analysis in small PCR systems and is expected to contribute to improving the performance of point-of-care diagnostics, thereby increasing accessibility to on-site diagnostics in the future.