Functional brain network identification and fMRI augmentation using a VAE-GAN framework

• Published in: Computers in biology and medicine Vol. 165; p. 107395
• Main Authors: Qiang, Ning, Gao, Jie, Dong, Qinglin, Yue, Huiji, Liang, Hongtao, Liu, Lili, Yu, Jingjing, Hu, Jing, Zhang, Shu, Ge, Bao, Sun, Yifei, Liu, Zhengliang, Liu, Tianming, Li, Jin, Song, Hujie, Zhao, Shijie
• Format: Journal Article
• Language: English
• Published: Oxford Elsevier Ltd 01.10.2023; Elsevier Limited
• Subjects: Attention deficit hyperactivity disorder; Brain; Brain disorders; Brain mapping; Coders; Data augmentation; Datasets; Deep learning; Design; fMRI; Functional brain network; Functional magnetic resonance imaging; Generative adversarial net; Identification; Machine learning; Magnetic resonance imaging; Modelling; Natural language; Neural networks; Neuroimaging; Performance enhancement; Random noise; Temporal variations; Variational auto-encoder; Variational auto-encoder; fMRI; Data augmentation; Functional brain network; Brain disorders; Generative adversarial net
• ISSN: 0010-4825; 1879-0534
• DOI: 10.1016/j.compbiomed.2023.107395

Recently, deep learning models have achieved superior performance for mapping functional brain networks from functional magnetic resonance imaging (fMRI) data compared with traditional methods. However, due to the lack of sufficient data and the high dimensionality of brain volume, deep learning models of fMRI tend to suffer from overfitting. In addition, existing methods rarely studied fMRI data augmentation and its application. To address these issues, we developed a VAE-GAN framework that combined a VAE (variational auto-encoder) with a GAN (generative adversarial net) for functional brain network identification and fMRI augmentation. As a generative model, the VAE-GAN models the distribution of fMRI so that it enables the extraction of more generalized features, and thus relieve the overfitting issue. The VAE-GAN is easier to train on fMRI than a standard GAN since it uses latent variables from VAE to generate fake data rather than relying on random noise that is used in a GAN, and it can generate higher quality of fake data than VAE since the discriminator can promote the training of the generator. In other words, the VAE-GAN inherits the advantages of VAE and GAN and avoids their limitations in modeling of fMRI data. Extensive experiments on task fMRI datasets from HCP have proved the effectiveness and superiority of the proposed VAE-GAN framework for identifying both temporal features and functional brain networks compared with existing models, and the quality of fake data is higher than those from VAE and GAN. The results on resting state fMRI of Attention Deficit Hyperactivity Disorder (ADHD)-200 dataset further demonstrated that the fake data generated by the VAE-GAN can help improve the performance of brain network modeling and ADHD classification. •The proposed VAE-GAN model inherits advantages of VAE (variational auto-encoder) and GAN (generative adversarial net) and avoids their limitations in modeling of fMRI data.•The VAE-GAN has superior performance in learning both temporal features and the corresponding functional brain networks from fMRI, and it can generate high-quality fake fMRI data.•The generated data from VAE-GAN can help improve the performance of brain network modeling and ADHD classification.