Fig. 2.

(a) Measured optical density and (b) transmittance of the four tapetal spectral filters in our bio-inspired multispectral imaging sensor. (c) Spatial uniformity or fixed pattern noise (FPN) before calibration as a function of light intensity. (d) The calibrated FPN shows that the bio-inspired camera captures data with ~0.1% spatial variations for uniform intensity targets. (e) Quantum efficiency of our bio-inspired image sensor. (f) Fluorescence detection limits under surgical light illumination for our bio-inspired sensor, which utilizes a multi-exposure method (green) compared with single-exposure method (blue) used in state-of-the-art pixelated color-NIR sensors. The dashed vertical lines are the individual detection thresholds for both sensors estimated as the mean value plus three standard deviations of the control vial. The bio-inspired sensor’s 1000 × improvement in detection limit is achieved due to multi-exposure (P < 0.0001), high NIR optical density, and high NIR quantum efficiency. Data are presented as mean ± SD.