Local similarity matching algorithm for determining SPR angle in surface plasmon resonance sensors

Surface plasmon resonance (SPR) sensors can detect the presence of molecules of interest at their surface, and can be used to determine their concentration, by comparing the SPR angle (or wavelength) for different samples before and after the reaction event. The degree of usefulness of such devices depends on the stability and sensitivity of the SPR angle determination. The stability means that the determined SPR angle is a function only of controllable parameters and the molecules of interest. The sensitivity means that determined SPR angles are distinct for measurements of distinct concentrations of the molecules of interest. We propose a new method, called “local similarity matching” (LSM), for a determination of the SPR angle that is stable. LSM is sufficiently immune to the uncontrollable fluctuations of input power and sensor noise to allow reliable discrimination of SPR angles differing by 0.01°. Experimental comparison shows that LSM is more stable than existing methods such as centroid, constant-reflectance, polynomial fit, and zero-point-derivative methods.