Spectral phasor analysis for brillouin microspectroscopy

Abstract

Brillouin Light Scattering (BLS) spectroscopy allows for the all-optical measurement of the hypersonic velocities in a sample, from which one can extract high-frequency elastic moduli. Recent advances in high resolution optical imaging spectrometers have made it conducive to studying live biological samples and further its implementation as an imaging modality for the life sciences. One major challenge in this context is the relatively weak BLS signal together with the subtle BLS spectral variations observed in many biological samples. Here we show that using spectral phasor analysis one can more easily distinguish variations in noisy spectra compared to standard least-squares (LS) fitting. There is no fitting in phasor analysis, and it is both robust in regards to unaccounted for functional variations in the spectra, and orders of magnitude faster than LS-fitting. As such it can prove particularly useful for increasing contrast in Brillouin imaging as well as for high-throughput BLS applications such as cell-sorting and medical diagnostics especially for statistically compromised or noisy data.