BEN: a generalizable Brain Extraction Net for multimodal MRI data from rodents, nonhuman primates, and humans

Extracting brain tissue is usually a vital first step in processing brain magnetic resonance imaging (MRI) data in neuroimaging studies. While several conventional tools have been optimized to handle human brain data, there have been no generalizable methods to extract brain tissues for multimodal MRI data from rodents, nonhuman primates, and humans. In the present study, we propose a domain-adaptive and semi-supervised deep neural network, named the Brain Extraction Net (BEN), to extract brain tissues across species, MRI modalities and platforms while requiring minimal or even no annotations. We have evaluated BEN on 18 independent datasets including 1,029 rodent and nonhuman primate MRI scans as well as 4,601 human MRI scans, covering five species, four modalities, and six platforms with various magnetic field strengths. Compared to conventional toolboxes, the superiority of BEN is illustrated by its accuracy, robustness, and generalizability. BEN provides uncertainty maps to quantify the confidence of the network's decisions and to model interrater variability. We design BEN as an open-source software that provides interfaces compatible with several popular neuroimaging toolboxes. Our results demonstrate that BEN has great promise to enhance neuroimaging studies at high throughput for both preclinical and clinical applications. ### Competing Interest Statement The authors have declared no competing interest.