Practical sensorless aberration estimation for 3D microscopy with deep learning

• DOI: 10.1364/oe.401933
• Publication Date: 2020-08-21T21:00:09Z
• Authors: Debayan Saha; Uwe Schmidt; Qinrong Zhang; Aurelien Barbotin; Qi Hu; Na Ji; Martin J. Booth; Martin Weigert; Eugene W. Myers
• Keywords: Python; Ground truth

Estimation of optical aberrations from volumetric intensity images is a key step in sensorless adaptive optics for 3D microscopy. Recent approaches based on deep learning promise accurate results at fast processing speeds. However, collecting ground truth microscopy data training the network typically very difficult or even impossible thereby limiting this approach practice. Here, we demonstrate that neural networks trained only simulated yield predictions real experimental images. We validate our and datasets acquired with two different modalities, also compare to non-learned methods. Additionally, study predictability individual respect their requirements find symmetry wavefront plays crucial role. Finally, make implementation freely available as open source software Python.