Model for the Analysis of Coherent Raman Spectra

To allow a thoughtful pre-processing of coherent Raman spectra (CRS) such as coherent anti-Stokes Raman spectra and stimulated Raman spectra and a comparison of spontaneous Raman and CRS measurement techniques, a modeling of these processes with special emphasize on the wavelength dependency was started within CRIMSON (See figure 1 for a visualization of CARS). This modeling will lead to an understanding of the contributions of different processes to the overall CRS/Raman signal.

The modeling is also relevant for artificial data generation needed to train deep learning techniques such as convolution neuronal networks (CNNs) and Long-short-term-memory networks (LSTMs) as phase retrieval tool. Based on the modeling we will standardize the wavelength/wavenumber description of the all spectral data.

Figure 1: CARS process is visualized, where pump (wavenumber= 0) photons and stokes photons (negative wavenumbers) will give rise to CARS photons (positive wavenumbers).