Optimized detector readout for biosensor

#The invention relates to a method of improved readout of an optical detector of biosensors and to a system to be used for said method.

#The demand for biosensors is increasingly growing these days. Usually, biosensors allow for the detection of a given specific molecule within an analyte, wherein the amount of said molecule is typically small. Therefore, target particles, for example super-paramagnetic label beads, are used which bind to a specific binding site or spot only, if the molecule to be detected is present within the analyte. One known technique to detect these label particles bound to the binding spots is FTIR. Therein, light is coupled into the sample at an angle of total internal reflection. If no particles are present close to the sample surface, the light is completely reflected. If, however, the label particles are bound to said surface, the condition of total internal reflection is violated, a portion of the light is scattered into the sample and thus the amount of light reflected by the surface is decreased. By measuring the intensity of the reflected light with an optical detector, it is possible to estimate the amount of particles bound to the surface.

#Typically, a photodiode is used as an optical detector. However, a detection by a (CCD) camera or any other multi-pixel system is much more efficient, since a camera allows for the parallel detection of various binding sites or spots. The disadvantage of the use of a camera, though, is that accurate measurements of the intensity of the binding spots is difficult due to gain- and offset errors in the read-out system.

#FIG. 1 schematically shows a camera picture obtained by FTIR.

#FIG. 2 illustrates a camera response for two different gains.

#FIG. 3 schematically shows an embodiment of an FTIR system according to the present invention.

#FIG. 1 schematically shows a camera picture obtained by FTIR. This picture is obtained by (predominantly) homogeneous illumination of the FTIR surface of a cartridge and projection of the reflected light via an optical system on a (CCD) camera. Therein, six binding sites or spots A1, A2 can be identified in a region B with a certain background illumination or noise level. The brightness of the binding sites depends on the amount of label particles bound to these sites: The more label particles are bound the more light is scattered at the respective region causing a decrease of the reflected light. Thus, a dark binding site indicates a large amount of bound label particles. The relative darkening D of a binding spot compared to the surrounding area is a measure for the number of bindings and accordingly for the amount of a specific molecule within the sample liquid.

#However, said relative darkening cannot be measured accurately because of gain- and offset errors in the read-out system. This is caused by unknown optical gain in the optical light path, the level of illumination, gain parameters in the camera (exposure time, shutter, active pixels etc.). Further negative effects are spurious background light and dark current in the camera.

#In a typical FTIR geometry, small signal changes from the label particles or beads are detected on a relatively large optical base line signal. Said optical base line signal originates from the large reflection from the surface of the binding surface. Due to this large optical base line signal, gain variations originating from temperature effects (drift) in the sensor, in the signal processing and optical light path will introduce large variations in the detected signal, which limits the achievable accuracy and detection limit of the biosensor. This is especially a problem during relatively long-time measurements for low target-concentrations.