Gradient Hyperalignment for Multi-subject fMRI Data Alignment
- 2.5k Downloads
Multi-subject fMRI data analysis is an interesting and challenging problem in human brain decoding studies. The inherent anatomical and functional variability across subjects make it necessary to do both anatomical and functional alignment before classification analysis. Besides, when it comes to big data, time complexity becomes a problem that cannot be ignored. This paper proposes Gradient Hyperalignment (Gradient-HA) as a gradient-based functional alignment method that is suitable for multi-subject fMRI datasets with large amounts of samples and voxels. The advantage of Gradient-HA is that it can solve independence and high dimension problems by using Independent Component Analysis (ICA) and Stochastic Gradient Ascent (SGA). Validation using multi-classification tasks on big data demonstrates that Gradient-HA method has less time complexity and better or comparable performance compared with other state-of-the-art functional alignment methods.
KeywordsHyperalignment Gradient ICA Brain decoding Classification
This work was supported in part by the National Natural Science Foundation of China (61473149, 61422204, and 61732006).
- 3.Yousefnezhad, M., Zhang, D.: Local discriminant hyperalignment for multi-subject fMRI data alignment. In: 34th AAAI Conference on Artificial Intelligence, pp. 59–61, San Francisco, USA (2017)Google Scholar
- 4.Chen, P.H., Chen, J., Yeshurun, Y., Hasson, U., Haxby, J.V., Ramadge, P.J.: A reduced-dimension fMRI shared response model. In: 28th Advances in Neural Information Processing Systems, Canada, pp. 460–468 (2015)Google Scholar
- 5.Chen, P.H., Guntupalli, J.S., Haxby, J.V., Ramadge, P.J.: Joint SVD-Hyperalignment for multi-subject FMRI data alignment. In: 24th IEEE International Workshop on Machine Learning for Signal Processing, France, pp. 1–6 (2014)Google Scholar
- 10.Xu, H., Lorbert, A., Ramadge, P.J., Guntupalli, J.S., Haxby, J.V.: Regularized hyperalignment of multi-set fMRI data. In: IEEE Statistical Signal Processing Workshop, USA, pp. 229–232 (2012)Google Scholar
- 12.Lorbert, A., Ramadge, P.J.: Kernel hyperalignment. In: 25th Advances in Neural Information Processing Systems, Harveys, pp. 1790–1798 (2012)Google Scholar
- 13.Chen, P.H., et al.: A convolutional autoencoder for multi-subject fMRI data aggregation. In: 29th Workshop of Representation Learning in Artificial and Biological Neural Networks, Barcelona (2016)Google Scholar
- 14.Zhang, H., Chen, P.H., et al.: A searchlight factor model approach for locating shared information in multi-subject fMRI analysis. arXiv preprint arXiv:1609.09432 (2016)
- 16.Yousefnezhad, M., Zhang, D.: Deep Hyperalignment. In: Conference on Neural Information Processing Systems, USA (2017)Google Scholar