During an assay, optical measurements are made at a sensing surface (112), and at least one “homogeneity-image” of the sensing surface (112) is generated. From this image, an “homogeneity-indicator” is determined for at least one region of interest, and the optical measurements are then evaluated in dependence on said indicator. The homogeneity-indicator may for example be a binary value which indicates if an inhomogeneity was detected or not. If an inhomogeneity was detected, all optical measurements may be rejected, only measurements for the involved region of interest may be rejected, or measurements for a selected sub-area of the involved region of interest (ROI) may be rejected.