High-throughput segmentation of unmyelinated axons by deep learning
Abstract Axonal characterizations of connectomes in healthy and disease phenotypes are surprisingly incomplete and biased because unmyelinated axons, the most prevalent type of fibers in the nervous system, have largely been ignored as their quantitative assessment quickly becomes unmanageable as th...
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2022-01-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-022-04854-3 |
| _version_ | 1819231484075048960 |
|---|---|
| author | Emanuele Plebani Natalia P. Biscola Leif A. Havton Bartek Rajwa Abida Sanjana Shemonti Deborah Jaffey Terry Powley Janet R. Keast Kun-Han Lu M. Murat Dundar |
| author_facet | Emanuele Plebani Natalia P. Biscola Leif A. Havton Bartek Rajwa Abida Sanjana Shemonti Deborah Jaffey Terry Powley Janet R. Keast Kun-Han Lu M. Murat Dundar |
| author_sort | Emanuele Plebani |
| collection | DOAJ |
| description | Abstract Axonal characterizations of connectomes in healthy and disease phenotypes are surprisingly incomplete and biased because unmyelinated axons, the most prevalent type of fibers in the nervous system, have largely been ignored as their quantitative assessment quickly becomes unmanageable as the number of axons increases. Herein, we introduce the first prototype of a high-throughput processing pipeline for automated segmentation of unmyelinated fibers. Our team has used transmission electron microscopy images of vagus and pelvic nerves in rats. All unmyelinated axons in these images are individually annotated and used as labeled data to train and validate a deep instance segmentation network. We investigate the effect of different training strategies on the overall segmentation accuracy of the network. We extensively validate the segmentation algorithm as a stand-alone segmentation tool as well as in an expert-in-the-loop hybrid segmentation setting with preliminary, albeit remarkably encouraging results. Our algorithm achieves an instance-level $$F_1$$ F 1 score of between 0.7 and 0.9 on various test images in the stand-alone mode and reduces expert annotation labor by 80% in the hybrid setting. We hope that this new high-throughput segmentation pipeline will enable quick and accurate characterization of unmyelinated fibers at scale and become instrumental in significantly advancing our understanding of connectomes in both the peripheral and the central nervous systems. |
| first_indexed | 2024-12-23T11:45:41Z |
| format | Article |
| id | doaj.art-eeebcd4fdbae4233a1b78aa47374606f |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-12-23T11:45:41Z |
| publishDate | 2022-01-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-eeebcd4fdbae4233a1b78aa47374606f2022-12-21T17:48:21ZengNature PortfolioScientific Reports2045-23222022-01-0112111610.1038/s41598-022-04854-3High-throughput segmentation of unmyelinated axons by deep learningEmanuele Plebani0Natalia P. Biscola1Leif A. Havton2Bartek Rajwa3Abida Sanjana Shemonti4Deborah Jaffey5Terry Powley6Janet R. Keast7Kun-Han Lu8M. Murat Dundar9Department of Computer and Information Sciences, Indiana University, Purdue UniversityDepartment of Neurology, Icahn School of Medicine at Mount SinaiDepartment of Neurology, Icahn School of Medicine at Mount SinaiBindley Bioscience Center, Purdue UniversityDepartment of Computer Science, Purdue UniversityDepartment of Psychological Sciences, Purdue UniversityDepartment of Psychological Sciences, Purdue UniversityDepartment of Anatomy and Physiology, The University of MelbourneWeldon School of Biomedical Engineering, Purdue UniversityDepartment of Computer and Information Sciences, Indiana University, Purdue UniversityAbstract Axonal characterizations of connectomes in healthy and disease phenotypes are surprisingly incomplete and biased because unmyelinated axons, the most prevalent type of fibers in the nervous system, have largely been ignored as their quantitative assessment quickly becomes unmanageable as the number of axons increases. Herein, we introduce the first prototype of a high-throughput processing pipeline for automated segmentation of unmyelinated fibers. Our team has used transmission electron microscopy images of vagus and pelvic nerves in rats. All unmyelinated axons in these images are individually annotated and used as labeled data to train and validate a deep instance segmentation network. We investigate the effect of different training strategies on the overall segmentation accuracy of the network. We extensively validate the segmentation algorithm as a stand-alone segmentation tool as well as in an expert-in-the-loop hybrid segmentation setting with preliminary, albeit remarkably encouraging results. Our algorithm achieves an instance-level $$F_1$$ F 1 score of between 0.7 and 0.9 on various test images in the stand-alone mode and reduces expert annotation labor by 80% in the hybrid setting. We hope that this new high-throughput segmentation pipeline will enable quick and accurate characterization of unmyelinated fibers at scale and become instrumental in significantly advancing our understanding of connectomes in both the peripheral and the central nervous systems.https://doi.org/10.1038/s41598-022-04854-3 |
| spellingShingle | Emanuele Plebani Natalia P. Biscola Leif A. Havton Bartek Rajwa Abida Sanjana Shemonti Deborah Jaffey Terry Powley Janet R. Keast Kun-Han Lu M. Murat Dundar High-throughput segmentation of unmyelinated axons by deep learning Scientific Reports |
| title | High-throughput segmentation of unmyelinated axons by deep learning |
| title_full | High-throughput segmentation of unmyelinated axons by deep learning |
| title_fullStr | High-throughput segmentation of unmyelinated axons by deep learning |
| title_full_unstemmed | High-throughput segmentation of unmyelinated axons by deep learning |
| title_short | High-throughput segmentation of unmyelinated axons by deep learning |
| title_sort | high throughput segmentation of unmyelinated axons by deep learning |
| url | https://doi.org/10.1038/s41598-022-04854-3 |
| work_keys_str_mv | AT emanueleplebani highthroughputsegmentationofunmyelinatedaxonsbydeeplearning AT nataliapbiscola highthroughputsegmentationofunmyelinatedaxonsbydeeplearning AT leifahavton highthroughputsegmentationofunmyelinatedaxonsbydeeplearning AT bartekrajwa highthroughputsegmentationofunmyelinatedaxonsbydeeplearning AT abidasanjanashemonti highthroughputsegmentationofunmyelinatedaxonsbydeeplearning AT deborahjaffey highthroughputsegmentationofunmyelinatedaxonsbydeeplearning AT terrypowley highthroughputsegmentationofunmyelinatedaxonsbydeeplearning AT janetrkeast highthroughputsegmentationofunmyelinatedaxonsbydeeplearning AT kunhanlu highthroughputsegmentationofunmyelinatedaxonsbydeeplearning AT mmuratdundar highthroughputsegmentationofunmyelinatedaxonsbydeeplearning |