Background: Conventional processing of nerve for histomorphometry is resource-intensive, precluding use in intraoperative assessment of nerve quality during nerve transfer procedures. Stimulated Raman scattering (SRS) microscopy is a label-free technique that enables rapid and high-resolution histology.
Methods: Segments of healthy murine sciatic nerve, healthy human obturator nerve, and human cross-facial nerve autografts were imaged on a custom SRS microscope. Myelinated axon quantification was performed through segmentation using a random forest machine learning algorithm in commercial software.
Results: High contrast, high-resolution imaging of nerve morphology was obtained with SRS imaging. Automated myelinated axon quantification from cross-sections of healthy human nerve imaged using SRS was achieved.
Conclusions: Herein, we demonstrate the use of a label-free technique for rapid imaging of murine and human peripheral nerve cryosections. We illustrate the potential of this technique to inform intraoperative decision-making through rapid automated quantification of myelinated axons using a machine learning algorithm.
Keywords: artificial intelligence; fluorescence; histology; nonlinear optical microscopy; peripheral nerve injuries.
© 2020 Wiley Periodicals, Inc.