Transfer Learning for Automated Anterograde Tracer Signal Segmentation in Marmoset Brain Microscopy Images
Keywords: Brain microscopy images, neuron tracer segmentation, transfer learning
Abstract: A goal of contemporary neuroscience research is to map the structural connectivity of primate brains using microscopy imaging data. In this context, the Japan Brain/MINDS project aims at using neural tracers to map neural connections in the marmoset brain. As part of the analysis, automated segmentation of tracer signals in microscopy images is demanded. We process two kinds of tracer images based on retrograde neural tracers and anterograde tracer images. While retrogradely connected cell bodies can be manually annotated quite easily in a reasonable time, the annotation of entire anterograde tracer images is impractical. We found that with transfer learning, a small amount of training data is sufficient to adapt networks that have been trained for the detection of retrogradely cell bodies to successfully segment cells and axonal projections in anterograde tracer images. In this paper, we explain the methodology and promising preliminary results.
Paper Type: validation/application paper
Primary Subject Area: Transfer Learning and Domain Adaptation
Secondary Subject Area: Segmentation
Paper Status: original work, not submitted yet
Source Code Url: The code will be publicly available after journal publication (2021/2022).
Data Set Url: The dataset will be publicly available after journal publication (2021/2022).
Registration: I acknowledge that publication of this at MIDL and in the proceedings requires at least one of the authors to register and present the work during the conference.
Authorship: I confirm that I am the author of this work and that it has not been submitted to another publication before.
TL;DR: Transfer learning scheme for anterograde tracer signal segmentation from retrograde cell bodies segmentation in marmoset brain microscopy images.
4 Replies
Loading