Abstract
Recent works on smart scanning techniques in Raman micro-imaging demonstrate the possibility of highly reducing acquisition time. In particular, Gilet et al. [Optics Express32, 932 (2024)10.1364/OE.509736] proposed a protocol combining compression in both spectral and spatial domains by focusing on essential information. This protocol consists of a two-pass scan in Raman modality at different signal-to-noise ratios (SNR). The first scan of the entire sample area at low SNR, and was identified as the bottleneck of the whole process. We propose revisiting this protocol by replacing this first scan with scanning electron microscopy (SEM), which is a faster imaging modality. We demonstrate that acquiring real data of biomedical interest according to this new protocol is three times faster, with limited distortion on the reconstructed Raman spectra and preserved clinical value of the extracted information. This is illustrated on bone samples for which SEM is correlated with Raman. We discuss the potential extension of this method to other slow spectral imaging modalities conventionally based on raster scans.