Techniques and applications for real-time spectral imaging: from gas sensing to in vivo imaging

The concept of spectral imaging is very simple: by recording an image in multiple spectral bands it is possible to extract additional information from subtle variations in spectral signatures that are invisible  in conventional images. The fundamental problems associated with the implementation and exploitation of spectral imaging are: how to record the three-dimensional data cube (two transverse dimensions and wavelength) in ‘real-time’ using a conventional two-dimensional detector and how to invert the recorded data to extract useful and reliable information. I will describe novel optical techniques that uniquely enable recording of the spectral data cube in a single snapshot, which can be inverted using physical models for light propagation in complex structures to yield quantitative chemical images. Real-world applications range from in vivo biophotonic imaging of oxygen consumption in biomedical research and ophthalmology to thermal infrared imaging of gas spectra.