U-net model for brain extraction: Trained on humans for transfer to non-human primates

Abstract

Brain extraction (a.k.a. skull stripping) is a fundamental step in the neuroimaging pipeline as it can affect the accuracy of downstream preprocess such as image registration, tissue classification, etc. Most brain extraction tools have been designed for and applied to human data and are often challenged by non-human primates (NHP) data. Amongst recent attempts to improve performance on NHP data, deep learning models appear to outperform the traditional tools. However, given the minimal sample size of most NHP studies and notable variations in data quality, the deep learning models are very rarely applied to multi-site samples in NHP imaging. To overcome this challenge, we used a transfer-learning framework that leverages a large human imaging dataset to pretrain a convolutional neural network (i.e. U-Net Model), and then transferred this to NHP data using a small NHP training sample. The resulting transfer-learning model converged faster and achieved more accurate performance than a similar U-Net Model trained exclusively on NHP samples. We improved the generalizability of the model by upgrading the transfer-learned model using additional training datasets from multiple research sites in the Primate Data-Exchange (PRIME-DE) consortium. Our final model outperformed brain extraction routines from popular MRI packages (AFNI, FSL, and FreeSurfer) across a heterogeneous sample from multiple sites in the PRIME-DE with less computational cost (20 s~10 min). We also demonstrated the transfer-learning process enables the macaque model to be updated for use with scans from chimpanzees, marmosets, and other mammals (e.g. pig). Our model, code, and the skull-stripped mask repository of 136 macaque monkeys are publicly available for unrestricted use by the neuroimaging community at https://github.com/HumanBrainED/NHP-BrainExtraction .

Transfer Learning Approaches for Neuroimaging Analysis: A Scoping Review

“…There are different strategies in transfer learning in terms of how the layers are transferred (e.g., directly, fine-tuned, or reinitialized) (Kalmady et al, 2021 ; Prakash et al, 2021 ; Ren et al, 2021 ; Wang et al, 2021 ; Weiss et al, 2021 ). For example, in CNN, there are convolution layers and fully-connected (FC) layers and in transfer learning one can transfer weights associated with convolution layers, FC layers, or both, then decide to freeze or fine-tune them.…”

mrHARDIflow : A pipeline tailored for the preprocessing and analysis of Multi-Resolution High Angular diffusion MRI and its application to a variability study of the PRIME-DE database

“…Brain masks were computed prior to processing on the anatomical T1 images using DeepBet v1.0 [Wang et al (2021)], a U-Net trained on macaque data from the PRIME-DE. All masks were quality controlled and manually fixed to prevent exclusion of brain voxels and inclusion of skull and eyes.…”

A Macaque Brain Extraction Model Based on U-Net Combined with Residual Structure

“…Humans and macaque are both primates, and the MRI signals from their brain tissue are similar. Using transfer learning, the model training for the macaque brain extraction method can achieve better performance [ 38 ]. However, due to the multicenter and limited sample characteristics of the macaque brain tissue data, mentioned above, the target domain’s feature distribution is unbalanced in applying transfer learning.…”