Accurate morphometric analysis is essential to studying nerve development, injury, and regeneration, but current approaches are inefficient, inconsistent, or inaccessible. Here, we present AutoSNAP, a deep learning framework for automated nerve morphometry assessment. AutoSNAP employs a custom convolutional neural network trained on both injured and non-injured murine nerve cross sections.
Segmentation performance is enhanced by the incorporation of confidence-weighted patch restitching and geometry-informed separation algorithms, both tailored specifically to nerve morphology. When compared to manual annotation, AutoSNAP demonstrated pixel-wise segmentation accuracy of 98.6% for myelin sheaths and 96.7% for axons. AutoSNAP also provides a comprehensive set of fiber-, axon-, and myelin-level morphometric characteristics.
These metrics showed strong agreement with manually obtained metrics, as evidenced by inter-class correlation coefficients and Pearson correlation coefficients > 0.9 for almost all metrics. The complete pipeline is deployed as a publicly accessible web application available at https://autosnap-7cia.onrender.com/. This work was supported in part by funding from the American Foundation for Surgery of the Hand, the American College of Surgeons Career Development Award, the OREF Mentored Clinician Scientist Award and a K08 Career Development Award from the NIH, K08 AR080260-01.
Department of Orthopaedic Surgery, Washington University School of Medicine in St. Louis, 660 South Euclid Ave, St. Louis, MO, 63110, USA Sundeep Chakladar, Owen Lee, Daniel A.
Hunter, Adam Boukind, Christopher J. Brogan The authors declare no competing interests. Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Below is the link to the electronic supplementary material. Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it.
Extract — continue reading at the source.
