biostudies-literatureOostrom MNational Institute of Neurological Disorders and StrokeNIMH NIH HHSNINDS NIH HHSNational Institute of Mental Healthe0293856https://www.ebi.ac.uk/biostudies/studies/S-EPMC10980229biostudies-literatureUnknown19(3)Light-sheet microscopy has made possible the 3D imaging of both fixed and live biological tissue, with samples as large as the entire mouse brain. However, segmentation and quantification of that data remains a time-consuming manual undertaking. Machine learning methods promise the possibility of automating this process. This study seeks to advance the performance of prior models through optimizing transfer learning. We fine-tuned the existing TrailMap model using expert-labeled data from noradrenergic axonal structures in the mouse brain. By changing the cross-entropy weights and using augmentation, we demonstrate a generally improved adjusted F1-score over using the originally trained TrailMap model within our test datasets.PloS oneFine-tuning TrailMap: The utility of transfer learning to improve the performance of deep learning in axon segmentation of light-sheet microscopy images.PMC10980229RF1 MH120119R01NS104944R01 NS104944R01 NS081071R01NS081071RF1 MH128969RF1MH120119RF1MH128969Wu ZBramer LMWebb-Robertson BJMMao TOostrom MMuniak MAWang WEichler West RMPande PBowyer KObiri MAkers SfalseFine-tuning TrailMap: The utility of transfer learning to improve the performance of deep learning in axon segmentation of light-sheet microscopy images.Light-sheet microscopy has made possible the 3D imaging of both fixed and live biological tissue, with samples as large as the entire mouse brain. However, segmentation and quantification of that data remains a time-consuming manual undertaking. Machine learning methods promise the possibility of automating this process. This study seeks to advance the performance of prior models through optimizing transfer learning. We fine-tuned the existing TrailMap model using expert-labeled data from noradrenergic axonal structures in the mouse brain. By changing the cross-entropy weights and using augmentation, we demonstrate a generally improved adjusted F1-score over using the originally trained TrailMap model within our test datasets.2024-01-01T00:00:00Z20242026-07-09T11:38:31.441Z2025-04-04T20:16:52.458ZS-EPMC109802293855193510.1371/journal.pone.0293856