Synthetic Data from Tractography Reduces Manual Annotation Needs for Fiber Bundle Segmentation in Primate Brain Histology
By
[Submitted on 25 Jun 2026]
Summary
This paper presents a synthetic-data augmented framework for automated fiber bundle segmentation in macaque tracer histology. The approach uses ex vivo diffusion MRI (dMRI) tractography as a generative prior to synthesize 2D image patches for training a U-Net model. By composing realistic foreground textures from tractography with backgrounds from blockface photos and applying domain randomization, the method achieves performance comparable to state-of-the-art while requiring 3x less manually annotated data. Experiments show improved generalization across brains and fiber bundle densities compared to training with real data only, though synthetic-only training performs poorly, highlighting the need for real supervision.
Source
bskySynthetic Data from Tractography Reduces Manual Annotation Needs for Fiber Bundle Segmentation in Primate Brain Histologyarxiv.orgKey quotes
· 5 pulledTracer injection studies in non-human primates provide a gold standard for validating dMRI tractography.
Our approach uses ex vivo dMRI tractography as a generative prior to synthesize 2D image patches for training.
Experiments on held-out brains demonstrate improved generalization across brains and fiber bundle densities compared to training with real data only.
Training with synthetic data only leads to poor performance, underscoring the need for real supervision.
Overall, our approach achieves performance comparable to the state-of-the-art while requiring 3x less manually annotated data.
You might also wanna read
Guide to Neuroimaging Datasets for Visual Perception Reconstruction from fMRI Data
This repository provides an index and overview of open neuroimaging datasets specifically for reconstructing visual perception from human fM
github.com·3mo ago
Training-Free Single-Image Diffusion Model Achieves Fast, High-Quality Generation
This paper presents a training-free approach to single-image diffusion models. Instead of training a neural network on a single image (which
arxiv.org·20d ago
Brain-IT: Transformer-Based Image Reconstruction from fMRI Brain Recordings
Brain-IT is a novel AI method that reconstructs images from fMRI brain recordings using a Brain-Interaction Transformer. The system achieves
amitzalcher.github.io·7mo ago
Nature research paper: A mosaic of whole-body representations on the human precentral gyrus
DDN: Discrete Distribution Networks - A Novel Generative Model for Image Reconstruction
DDN (Discrete Distribution Networks) is a novel generative model accepted by ICLR 2025 that introduces a unique approach to image generation
discrete-distribution-networks.github.io·8mo ago
Comments
Sign in to join the conversation.
No comments yet. Be the first.