Deep learning spectroscopic stimulated Raman scattering microscopy

摘要

Spectroscopic stimulated Raman scattering (SRS) is a label-free chemical imaging modality enabling visualizationof molecules in living systems with high specificity. Among various spectroscopic SRS imaging methods, aconvenient way is through linearly chirping two femtosecond lasers and tuning their temporal delay, which inturn corresponds to different Raman shifts. Currently, the acquisition speed using a resonant mirror is 3 seconds(80 microseconds per spectrum), which is insuu000ecient for imaging samples with high motility. In this work, weaim to push the imaging speed using a 50-kHz polygon scanner as a delay line tuner, achieving a speed of 20microseconds per spectrum. At such high speeds, to overcome the signal level decrease due to reduced signalintegration time, we apply a U-Net deep learning framework, which first takes pairs of spectroscopic SRS imagesat different speeds as training samples, with high-speed, low-signal images as input and low speed, high-signalones as output. After training, the network is capable of rapidly transforming a low-signal spectroscopic image toa high-signal version. Consequently, our design can generate ultrafast spectroscopic SRS image while maintainingthe signal level comparable to the output with longer signal integration time.