An Innovative Machine Learning Based Multistage Signal Amplification Method Breaks through the Detection Limits of Conventional Optical Sensors

Surface plasmon resonance (SPR) enables in situ, label-free, real-time molecular detection. However, traditional SPR biosensors require complex and precise large-scale equipment, with detection limits constrained by multiple factors, complicating and prolonging biosensor development, especially in emergencies. To address these challenges, we introduce an innovative multistage signal amplification metasurface plasmon resonance (MSA-MetaSPR) method based on a complete inverse design. Adherent to the principle of signal maximization, we divided the system into three main components: the sensor, the capturer, and the signal acquisition unit. By introducing an optical density (OD)-based inverse design method, we designed a MetaSPR chip with a sensitivity of up to 573 nm/RIU and developed artificial antibodies with a 3-fold increase in affinity. Finally, we introduced a novel analytical method for processing biosensor data. This complete inverse design-based multistage amplification method achieves a nearly 1200-fold improvement compared to undesign-based sensors and an almost 150-fold improvement over conventional SPR methods. This proposed approach accelerates the development of biosensors for urgent situations and significantly advances the capabilities of SPR sensor technologies.